HANDBOOK

2015 AGTA Conference

CONTENTS

WELCOME 5

GENERAL INFORMATION 7

AGTA15 CONFERENCE PROGRAM 10

CONFERENCE SOCIAL PROGRAM 19

ABSTRACTS & BIOGRAPHIES 21

POSTER PRESENTATIONS MONDAY 62

POSTER PRESENTATIONS TUESDAY 77

SPONSORS 96

EXHIBITORS 98

AGTA 2015 DELEGATE LIST 103

EXHIBITOR FLOOR PLAN 108

CONFERENCE MANAGERS

Leishman Associates 113 Harrington Street, Hobart TAS 7000 170 Elgin Street, Carlton VIC 3053 P. 03 6234 7844 F. 03 6234 5958 E. [email protected] W. www.leishman-associates.com.au

3 2015 AGTA Conference

AGTA AGTA 2015 Conference Executive Team Organising Committee Ruby C Y Lin (President), Asbestos Diseases Marcel Dinger, Garvan Institute of Medical Research Institute Research (co-convenor)

Carsten Kulheim (Vice President), Australian Carsten Kulheim, The Australian National National University University (co-convenor)

Mark van der Hoek (Treasurer), South Australian Jonathan Arthur, University of Sydney Health and Medical Research Institute Nikola Bowden, University of Newcastle Vikki Marshall (Secretary), Melbourne Neuroscience Institute, University of Melbourne Daniel Catchpoole, The Children’s Hospital at Westmead Alicia Oshlack, Murdoch Children’s Research Institute Aaron Darling, University Of Technology Sydney

Richard Tothill, Peter MacCallum Cancer Centre Gyorgy Hutvagner, University of Technology Sydney Erik (Rik) Thompson (Founding AMATA President) Ruby Lin, Asbestos Diseases Research Institute

Jac Charlesworth, Menzies Institute for Medical Vikki Marshall, University of Melbourne Research Pablo Moscato, University of Newcastle Daniel Catchpoole, The Children’s Hospital at , Murdoch Children’s Research Westmead Alicia Oshlack Institute Ryan Lister, The University of Western Australia Ian Paulsen, Macquarie University Ian Paulsen, Macquarie University Helen Speirs, Ramaciotti Centre For Genomics Marcel Dinger, Garvan Institute of Medical , Peter MacCallum Cancer Centre Research Richard Tothill Torsten Thomas, University of New South Wales Liam Williams, Auckland University Jean Yang, University Of Sydney Mark Waltham, University of Melbourne

Kirby Siemering, Australian Genome Research Facility

Mark Crowe, QFAB Bioinformatics

Andreas Schreiber, Centre for Cancer Biology

Nicole Cloonan, QIMR Berghofer Medical Research Institute

Robert Day, CTCR University of Otago

4 2015 AGTA Conference

their time, contribution and support on the AGTA WELCOME managing executive committee and I am happy to FROM THE AGTA say that they remain firm supporters of AGTA. This year, it is only befitting that we have PRESIDENT a theme on genomics of plants and fine wine, being in the Hunter Valley. We also have focussed themes on transcriptomics, Dear Delegates and developmental genomics, cancer genomics, Invited Guests, clinical genomics, quantitative genetics, microbial and single cell genomics. On behalf On behalf of the of the managing executives and the local Australasian Genomic conference organising committee, I would like Technologies Association to welcome our international keynote speakers managing executives, I to Hunter Valley. We hope you enjoy and take welcome you to the 15th Annual Scientific Meeting the time to explore our beautiful country. We are for AGTA in the heart of the wine country, Hunter also thankful to our invited speakers and national Valley, New South Wales, Australia. chairs. AGTA 2015 committee is very proud to showcase many early career scientists including This meeting marks the first year of the students, who are AGTA members. I would like association under the new banner of AGTA and to thank each one of you for contributing to this we remain the premium platform to connect with dynamic line-up and I hope that you enjoy the many of Australia’s leading scientists in the field of program ahead. genomics. This is where people from academia, industry, research and the clinic converge to have We are very thankful for the participation of our rigorous discussion. For the newcomers to the Gold Sponsor BD Biosciences, Silver Sponsors; Association, this is where you form long-term Roche Diagnostics and Illumina, and Qiagen friendships, networks and collaborations, and find and DNAnexus for sponsored speakers. We are career opportunities. also thankful for some twenty exhibitors, without whom this meeting would not be possible. We We have had a busy year consolidating our recognise the importance technology companies infrastructure, governance and communication play in the evolution of genomics in Australia with our members. We have finalised and and we strongly encourage you to take the time registered our new constitution as AGTA. Our and opportunity to interact with these leading society continues with the Small Grant Scheme, companies. which resulted in two successful recipients, announced in June 2015. We continue our It is my pleasure to welcome you to AGTA 2015 collaboration with QFAB who host our website, on behalf of the convenors, Marcel Dinger and ww.agtagenomics.org.au. We have revamped Carsten Külheim, the AGTA 2015 conference the news section to include job postings and organising committee and the AGTA managing introduced notifications of workshops in AGTA executives. We trust that you will enjoy AGTA members’ respective states. To increase 2015. communication with our members, we are active on Facebook, Twitter and LinkedIn and we have introduced a 3-year membership category upon members’ feedback. We hope these interactions Dr Ruby CY Lin and support continue as our membership grows President into the future. We are proud to maintain AGTA Australasian Genomic Technologies Association meetings as gender balanced and family friendly. This year we farewell our executives, Mark Crowe, Daniel Catchpoole and Ian Paulsen. I would like to take this opportunity to thank them for

5 2015 AGTA Conference

transcriptomics, small and non-coding RNAs, WELCOME FROM disease, data handling and analysis, population THE CONFERENCE genetics and epigenomics. Along with the excellent scientific program we CONVENORS hope you will enjoy the scenery and hospitality of the Hunter Valley. The AGTA (formerly AMATA) meetings have been the birthplace for many friendships, collaborations and good memories. We trust that this year’s meeting will continue this tradition and remind us that some of Australasia’s best science has been the product of strong scientific networks.

We gratefully acknowledge the invited Dear Delegates and Invited Guests, international, national and guest speakers for their time, energy, and invaluable contribution to On behalf of the AGTA Executive Committee AGTA 2015, and to our sponsors for their support and the 2015 AGTA Conference Organising without which this conference could not take Committee, we extend to you a very warm place. welcome to the Hunter Valley and the 15th Annual Conference of the Australasian Genomic Once again welcome, thank you for your Technologies Association. contribution, and please savour the insights good science, and maybe the odd wine, brings. New technologies and innovations are reshaping scientific research at unprecedented rates - enabling new approaches to solve biological problems that could not be foreseen even a few Associate Professor Marcel Dinger years ago. What better place to appreciate the Garvan Institute of Medical Research biological insights derived from next-generation sequencing and related technologies than the Dr Carsten Külheim Hunter Valley, especially when they are related to Australian National University wine.

The organising committee has been overwhelmed by the very high calibre of research presented in submitted abstracts, making it a challenge to select a limited number of oral presentations. However, the committee has managed to compile what we consider to be an exciting and broad scientific program that encompasses

6 2015 AGTA Conference

General information Conference Name Badges All delegates, speakers, sponsors and exhibitors Registration Desk will be provided with a name badge, which must Please direct any questions you may have be worn at all times within the conference venue, regarding registration, accommodation or social as it is required for access to all the conference functions to Leishman Associate staff at this desk. sessions and social functions.

Registration Desk Opening Times With thanks to our Name Badge Sponsor:

Sunday 11 October 2015 2:00pm - 8:00pm

Monday 12 October 2015 Conference Proceedings 7:30am - 5:30pm Speaker PowerPoints and abstracts will be Tuesday 13 October 2015 available on the AGTA website following the 7:30am - 5:30pm conclusion of the conference. Speakers will be requested to sign a release form. This is not Wednesday 14 October 2015 compulsory. 8:00am - 2:00pm Conference WIFI Accommodation Wireless internet will be available throughout If you have any queries relating to your the conference venue for the duration of accommodation booking first speak to the staff at the conference. A username and password your hotel or alternatively Leishman Associates will be given to you when you register at the staff at the Registration Desk. conference. Your credit card details were supplied to the hotel Dress Codes you have selected, as security for your booking. If you have arrived 24 hours later than your Dress throughout the day is smart casual or indicated arrival day you may find that you have informal business. been charged a fee. You will be responsible for all room and incidental charges on check out and Emergency Medical Care may be asked for an impression of your credit card for security against these charges. This is For any medical emergency please telephone standard policy in many hotels. 000. The staff at your hotel will have information if you require contact details for a doctor, dentist or other health professional.

EXHIBITOR PRIZE DRAW

An exhibitor passport will be given to all delegates at registration. The AGTA 2015 organising committee encourages you to visit each trade exhibitor and have your passport stamped, to go into the draw to win some great prizes!

7 2015 AGTA Conference

Social Program Entry their session chair and to familiarise themselves with the room and the audio visual equipment. The Welcome Reception is included in the cost For information on the chairperson attending your of each full conference registration. session, please see the Registration Desk.

With thanks to our Welcome Reception Sponsor: A technician will be present in the speaker’s preparation room during registration hours. There will be facility to test and modify your presentation as required.

Oral Presentations The Conference Dinner IS NOT included in any registration type. Social events ARE NOT included Please refer to the program for the time allocated in the cost of day registrations or for accompany- for each presentation, as these do vary. The ing partners. Places for day registrants and addi- chairperson for your session will give you a tional guests for these events may still be available 3 minute warning, however you are asked to at an additional cost. Bookings can be made at the stick to your time allocation so that the program registration desk subject to availability. remains on schedule.

All delegates who are registered to attend the Poster Presentations dinner will receive a named sticker at registration. Posters will be displayed in the Exhibition You MUST place your sticker on a table located Centre of the Crowne Plaza for the duration of on poster boards next to the registration desk. the conference. There will be a poster session You must allocate yourself to a table no later than on Monday 12 October 2015 from 1.30pm to 11.00 am Tuesday 13 October 2015. 3.30pm and on Tuesday 13 October at 1.05pm to 3.00pm. STUDENT FUNCTIONS

All conference students and early career Special Diets researchers are invited to a casual function on All catering venues have been advised of any Monday 12 October. Please forward enquiries to special diet preferences you have indicated on the registration desk. Further information about your registration form. Please identify yourself this event can be found on page 19. to venue staff as they come to serve you and they will be pleased to provide you with all pre- Photographs, Videos, Recording of ordered food. For day catering, there may be a Sessions specific area where special food is brought out, Delegates are not permitted to use any type of please check with catering staff. camera or recording device at any of the sessions unless written permission has been obtained from SECURITY the relevant speaker. The members of the conference organising committee, Leishman Associates and The Speakers and Speaker’s Preparation Crowe Plaza Hunter Valley accept no liability for Room personal accident or loss or damage suffered by All speakers should present themselves to the any participant, accompanying person, invited Speaker’s Preparation Room, located next to observer or any other person by whatever the conference registration desk at least 4 hours means. Nor do we accept liability for any before their scheduled presentation time, to upload equipment or software brought to the conference their presentation. by delegates, speakers, sponsors or any other party. Speakers are requested to assemble in their session room 5 minutes before the commencement of their session, to meet with

8 2015 AGTA Conference

Please protect your personal property. Do not Disclaimer leave laptops, cameras, and other valuable items unsecured. Be conscious of individuals The 2015 AGTA Conference reserves the right who appear out of place and do not wear a to amend or alter any advertised details relating conference name badge. Advise Leishman to dates, program and speakers if necessary, Associates staff if this does occur. without notice, as a result of circumstances beyond their control. All attempts have been Parents Room made to keep any changes to an absolute minimum. A private room will be available at the conference venue for nursing mothers and others with sensitive personal health needs. Please note that this room will not be staffed. Attendees are not permitted to utilise this room for babysitting services. Please see the Leishman Associates staff at the registration desk for further information

9 CONFERENCE PROGRAM

SUNDAY 11 October 2015

PRE-CONFERENCE WORKSHOP Cellar Room 2 1130 - 1630 GENOTYPING-BY-SEQUENCING: A LOW COST TOOL FOR POPULATION GENETICS WORKSHOP Mr Aaron Chuah, The Australian National University, ACT, Australia Mr Rob Elshire, The Elshire Group Limited, New Zealand Professor Justin Borevitz, The Australian National University, ACT, Australia 1400 - 2000 Conference Registration Open Galleria OPENING ORATION Sauvignon Room CHAIR: DR RUBY CY LIN

1700 - 1800 STUDYING CANCER AT SINGLE NUCLEOTIDE RESOLUTION Professor Sean Grimmond Wolfson Wohl Cancer Research Centre, Institute for Cancer Sciences, University of Glasgow, Scotland, U.K

1800 - 2000 Welcome Reception & Trade Exhibition Exhibition Centre Proudly sponsored by

2000 Free evening for delegates

Early Career Invited Guest Speaker Student Keynote Researcher Speaker

10 CONFERENCE PROGRAM

MONDAY 12 October 2015

0730 - 1730 Registration Desk Open & Welcome Refreshments Galleria 0800 - 1730 Trade Exhibition Open Exhibition Centre 0830 - 0845 O f fi c i a l W e l c o m e a n d C o n f e r e n c e O p e n i n g S a u v i g n o n R o o m SESSION 1: TRANSCRIPTOMICS AND THE REGULATION OF THE GENOME CHAIRS: DR PHILLIPPA TABERLAY & DR TIMOTHY MERCER Sauvignon Room 0845 - 0930 DIFFERENTIAL ANALYSIS OF BIFURCATING SINGLE-CELL GENE EXPRESSION TRAJECTORIES Assistant Professor Cole Trapnell University of Washington, U.S.A 0930 - 0955 INSIGHTS INTO MAMMALIAN CELL TYPES Professor Alistair Forrest Harry Perkins Institute of Medical Research, WA, Australia

0955 - 1015 ALTERATIONS IN THREE-DIMENSIONAL ORGANISATION OCCUR COINCIDENT WITH LONG-RANGE EPIGENETIC DEREGULATION OF THE CANCER GENOME Dr Phillippa Taberlay Garvan Institute of Medical Research, NSW, Australia 1015 - 1030 3’ÚTR DYNAMICS PREDICT mRNA FATE Dr Traude Beilharz Monash University, VIC, Australia

1030 - 1045 DYNAMIC EXPRESSION OF LONG NONCODING RNAs AND REPEAT ELEMENTS IN SYNAPTIC PLASTICITY Mr Jesper Maag Garvan Institute of Medical Research, NSW, Australia 1045 - 1115 Morning Refreshments & Trade Exhibition Exhibition Centre SESSION 2: GENOME EVOLUTION AND BIOINFORMATICS Sauvignon Room CHAIRS: DR NATALIE THORNE & DR JOSEPH POWELL 1115 - 1200 WHAT WE HAVE LEARNED FROM SEQUENCING ARCHAIC HUMAN GENOMES Dr Janet Kelso Max Planck Institute for Evolutionary Anthropology, Germany 1200 - 1225 USING GENE CO-EXPRESSION NETWORKS TO DISCOVER EPILEPSY GENES Associate Professor Melanie Bahlo Walter + Eliza Hall Institute of Medical Research, VIC, Australia

11 CONFERENCE PROGRAM

MONDAY 12 October 2015

1225 - 1245 DEEP GENE EXPLORATION OF HUMAN CHROMOSOME 21 Dr Timothy Mercer Garvan Institute of Medical Research, NSW, Australia

1245 - 1300 MAPPING STEM CELL BIOLOGY BY HIGH RESOLUTION TEMPORAL TRANSCRIPTOMICS Dr Brian Gloss Garvan Institute of Medical Research, NSW, Australia 1300 - 1400 Lunch & Trade Exhibition Exhibition Centre 1330 - 1530 Poster Session One & Afternoon Refreshments Exhibition Centre SESSION 3: DEVELOPMENTAL GENOMICS AND THE ORIGINS OF DISEASE CHAIR: PROFESSOR ANDREW PERKINS & DR KATE HOWELL Sauvignon Room 1530 - 1555 HIGHLY CONSERVED EPIGENOME REMODELLING DURING THE VERTEBRATE PHYLOTYPIC PERIOD Professor Ryan Lister The University of Western Australia, WA, Australia

1555 - 1615 GENOME-DRIVEN INSIGHTS TO DISEASE MECHANISMS Associate Professor Robyn Jamieson Children’s Medical Research Institute, NSW, Australia

1615 - 1635 ADVANCES IN THE UNDERSTANDING OF MUTATIONAL SIGNATURES IN HUMAN SOMATIC CELLS Dr Mirana Ramialison Australian Regenerative Medicine Institute, Monash University, VIC, Australia

1635 - 1650 REPLI-G SINGLE CELL SEQUENCING: DRIVING THE NEXT FRONTIER OF GENOMIC AND TRANSCRIPTOMIC ANALYSIS Mr Colin Baron Qiagen Life Sciences, U.S.A Proudly supported by

1650 - 1705 UNCOVERING HIDDEN GENES IN INTERGENIC GWAS REGIONS Dr Nenad Bartonicek Garvan Institute of Medical Research, NSW, Australia

12 CONFERENCE PROGRAM

MONDAY 12 October 2015

1705 - 1720 VISUALISATION OF VARIANTS USING AQUARIA: NON-SYNONYMOUS SNPs IN THE CONTEXT OF PROTEIN STRUCTURE Dr Neil Saunders Digital Productivity Flagship, CSIRO, NSW, Australia 1720 Free evening for delegates 1800 onwards Student Function The Lovedale Bar

Students and early career researchers are invited to join fellow delegates for a fun night of genomics trivia. Finger food and limited beverage service will be provided.

Please confirm your attendance to this function by advising the staff at the conference registration desk.

DON’T FORGET to visit the Trade Exhibitors and have your Passport stamped! You could win a $200 JB HI-FI Voucher

13 CONFERENCE PROGRAM

TUESDAY 13 October 2015

0730 - 1730 Registration Desk Open & Welcome Refreshments Galleria 0800 - 1730 Trade Exhibition Open Exhibition Centre 0825 - 0830 Welcome to Day Two Sauvignon Room SESSION 4: CANCER GENOMICS AND PRECISION MEDICINE Sauvignon Room CHAIRS: PROFESSOR DAVID THOMAS & DR NIC WADDELL 0830 - 0910 ADVANCES IN THE UNDERSTANDING OF MUTATIONAL SIGNATURES IN HUMAN SOMATIC CELLS Dr Serena Nik-Zainal Wellcome Trust Sanger Institute, Cambridge, U.K

0910 - 0945 DEVELOPMENTAL ANOMALIES AND PEDIATRIC CANCER: A GROWING CONNECTION Dr Todd Druley Washington University School of Medicine, U.S.A 0945 - 1015 HIGHLY PARALLEL MEASUREMENT OF THE IMPACT OF MUTATIONS IN PROTEINS Assistant Professor Douglas Fowler Department of Genome Sciences, University of Washington, U.S.A 1015 - 1030 GENOMIC CATASTROPHES DRIVING TUMORIGENESIS IN ESOPHAGEAL ADENOCARCINOMA Dr Katia Nones QIMR Berghofer, QLD, Australia 1030 - 1045 THE HISTONE H3 LYSINE 4 PRESENTER WDR5 IS A BIOMARKER IN N-MYC-INDUCED TRANSCRIPTIONAL ACTIVATION AND NEUROBLASTOMA PROGRESSION Dr Yuting Sun Children’s Cancer Institute, NSW, Australia 1045 - 1115 Morning Refreshments & Trade Exhibition Exhibition Centre

SESSION 5: CLINICAL GENOMICS AND THE FUTURE OF HEALTHCARE CHAIRS: DR CAS SIMONS & DR DENIS BAUER Sauvignon Room 1115 - 1155 FROM GENES TO GENOMES IN MEDICAL RESEARCH AND PATIENT CARE Professor Joris Veltman Radboud University Medical Centre and Maastricht University Medical Centre, The Netherlands

14 CONFERENCE PROGRAM

TUESDAY 13 October 2015

1155 - 1220 WHOLE GENOME SEQUENCING BASED CLINICAL GENOMICS Dr Mark Cowley Garvan Institute of Medical Research, NSW, Australia Proudly supported by

1220 - 1235 STRUCTURAL VARIATION CALLING FROM HiSeq X WHOLE GENOME SEQUENCING DATA IN A CLINICAL SETTING Dr Andre Minoche Garvan Institute of Medical Research, NSW, Australia 1235 - 1250 SHARED ANALYSIS FOR GENOME TESTING IN THE HEALTH CARE SETTING Dr Natalie Thorne Melbourne Genomics Health Alliance, VIC, Australia

1250 - 1305 BRAIN-cX – AN INTERACTIVE WEB-TOOL FOR GENE PRIORITISATION AND MORE Dr Saskia Freytag The Walter + Eliza Hall Institute of Medical Research, VIC, Australia 1305 - 1500 Lunch, Poster Session Two & Trade Exhibition Exhibition Centre 1430 - 1500 Afternoon Refreshments Exhibition Centre SESSION 6: GENOMICS OF PLANTS AND FINE WINE Sauvignon Room CHAIRS: PROFESSOR JUSTIN BOREVITZ & DR ROSE ANDREW 1500 - 1515 INTRODUCTION TO WINE TASTING BY GUNDOG ESTATE

1515 - 1555 ORIGIN AND CONSEQUENCES OF GENETIC AND EPIGENETIC VARIATION IN ARABIDOPSIS THALIANA AND ITS RELATIVES Professor Detlef Weigel Max Planck Institute for Developmental Biology, Germany 1555 - 1630 DEMOCRATISING GENETIC ANALYSIS AND BREEDING WITH GENOTYPING BY SEQUENCING Mr Rob Elshire The Elshire Group Limited, New Zealand

1630 - 1645 UNRAVELLING THE MāNUKA GENOME: A METAGENOMIC APPROACH Ms Amali Thrimawithana The New Zealand Institute for Plant & Food Research Limited, New Zealand

15 CONFERENCE PROGRAM

TUESDAY 13 October 2015

1645 - 1700 THE SMRT WAY TO SEQUENCE A YEAST GENOME Dr Richard Edwards University of New South Wales, NSW, Australia 1700 - 1730 WINE ‘OMICS: AT THE CUTTING-EDGE OF THE OLDEST BIOTECHNOLOGY Dr Anthony Borneman Australian Wine Research Institute, SA, Australia

1900 - 2300 Conference Dinner Lindemans Hunter Valley

Optional function at $130.00 per person. Bookings are essential.

Please assemble in the Crowne Plaza Foyer for a 1830 coach departure to Lindemans.

Coaches will operate a return shuttle to Crowne Plaza from 2145.

16 CONFERENCE PROGRAM

WEDNESDAY 14 October 2015

0800 - 1400 Registration Desk Open & Welcome Refreshments Galleria 0800 - 1330 Trade Exhibition Open Exhibition Centre 0855 - 0900 Welcome to Day Three Sauvignon Room SESSION 7: QUANTITATIVE GENETICS AND DECODING THE GENOME CHAIR: DR HELEN SPEIRS & DR ALEXANDRA LIVERNOIS Sauvignon Room 0900 - 0930 CAN WE EXPLORE THE GENOMICS OF CANINE WORKING BEHAVIOUR WITH SELECTIVE SWEEP ANALYSIS? Professor Claire Wade Faculty of Veterinary Science, The University of Sydney, NSW, Australia 0930 - 0945 PROGRAMMABLE RNA TARGETING AND CLEAVAGE BY CRISPR/CAS9 Dr Mitchell O’Connell Doudna Lab, Department of Molecular and Cell Biology, University of California, U.S.A 0945 - 1000 EXPRESSION QUANTITATIVE TRAIT LOCI (eQTLs) OF ENDOMETRIOSIS RISK LOCUS AT 1p36.12 Dr Jenny Fung QIMR Berghofer Medical Research Institute, QLD, Australia 1000 - 1045 Morning Refreshments & Trade Exhibition Exhibition Centre

AGTA Annual General Meeting Sauvignon Room

SESSION 8: MICROBIAL AND SINGLE CELL GENOMICS CHAIRS: DR KIRBY SIEMERING & DR CLARE STIRZAKER Sauvignon Room 1045 - 1130 ILLUMINATING MICROBIAL DARK MATTER VIA SINGLE-CELL GENOMICS Dr Christian Rinke University of Queensland, QLD, Australia 1130 - 1150 HIGH THROUGHPUT CELL SORTING FOR DOWNSTREAM GENOMIC ANALYSIS Dr J Clark Mason BD Biosciences, U.S.A Proudly supported by

17 CONFERENCE PROGRAM

WEDNESDAY 14 October 2015

1150 - 1205 SINGLE-CELL ANALYSIS OF BREAST CANCER MOLECULAR SUBTYPE REVEALS CLINICALLY RELEVANT HETEROGENEITY Ms Laura Baker Garvan Institute of Medical Research, NSW, Australia

1205 - 1220 THE EXTREME MICROBIOME PROJECT (XMP): THE METAGENOMICS OF LAKE HILLIER, A PINK HYPERSALINE LAKE Dr Ken McGrath The Australian Genome Research Facility, QLD, Australia 1220 - 1235 Awarding of Prizes, Conference Close and 2016 Conference Launch 1235 - 1315 Lunch & Trade Exhibition Exhibition Centre 1330 Coach Transfer departs Crowne Plaza to Sydney Domestic Airport

Please assemble in the Crowe Plaza Foyer by 1320 for a prompt coach departure at 1330.

18 2015 AGTA Conference

CONFERENCE SOCIAL PROGRAM Welcome Reception We will provide finger food and your first few drinks. Date: Sunday 11 October 2015 Venue: Crowne Plaza Hunter Valley, Please see Brigitte or Annika at the Exhibition Centre conference registration desk to register for Time: 6.00pm – 8.00pm this function. Dress: Smart Casual Conference Dinner Join us for the official Welcome Reception for the AGTA 2015 Conference. Enjoy networking Date: Tuesday 13 October 2015 with old and new acquaintances, and Venue: Lindemans Winery familiarising yourself with the trade exhibitors, Time: 7.00pm – 11.00pm whilst enjoying drinks and canapés. Dress: Smart Casual Attendance: By booking and paying prior to The Welcome Reception is included in a full the conference. You are still able to book at registration only. Additional tickets can be the conference, however it will be subject to purchased at $70.00 per person. availability.

Proudly supported by: The conference dinner is the social highlight of the conference and should not be missed.

Come and join us for another chance to network and meet with colleagues, whilst enjoying a great night of food, wine and dancing.

Student Function IMPORTANT INFORMATION Date: Monday 12 October 2015 Seating and table allocation for the AGTA Venue: Crowne Plaza Hunter Valley, Dinner will be by way of sticker allocation. The Lovedale Bar All delegates registered to attend the AGTA Time: 1830 onwards Dinner will receive a sticker to be placed on the table sheets near the Registration Desk. Students and early career researchers are These sheets will be available from Monday invited to join fellow delegates for a fun night 12 October and will be taken down at the of genomics trivia. Test your knowledge of end of morning refreshments on Tuesday 13 genomics technologies, bioinformatics tools, October or as they become full. the history of sequencing and much more! As in real life genomics, your team will have If you do not have a sticker please see the a diverse range of backgrounds, so will have Registration Desk staff, DO NOT write your to work together to win the prize! Teams will name directly on the board, as you will NOT form on the night, and points will be issued be allocated a seat. for both correct and creative answers.

19 Historical Breakthroughs

Breathtaking Progress Radical Reinvention

The NextSeq™ A Whole Human Genome on Your Desktop

© 2015 Illumina, Inc. All rights reserved. Images courtesy of Nature Publishing Group and Corbis Images. www.illumina.com/nextseq 2015 AGTA Conference

ABSTRACTS AND BIOGRAPHIES 2015 AGTA Conference

PRE-CONFERENCE WORKSHOP

GENOTYPING-BY-SEQUENCING: A LOW COST TOOL FOR POPULATION GENETICS WORKSHOP

Mr Aaron Chuah, The Australian National University, ACT, Australia Mr Rob Elshire, The Elshire Group Limited, New Zealand Professor Justin Borevitz, The Australian National University, ACT, Australia

Date: Sunday 11 October 2015 Venue: Cellar Room 2, Crowne Plaza Hunter Valley Time: 11.30am – 4.30pm

This is an optional workshop at $125.00 per person. Bookings are essential.

Genotyping-by-sequencing (GBS) can be performed cheaply and with a large number of samples, with or without a reference sequence, enabling a wide range of environmental, conservation and population genetics projects. Our target audience are wet-lab biologists who wish to perform their analyses in-house. We will be covering the lab preparation and basic bioinformatic analysis of this exciting new technology. If attendees wish to try their own hand at analysing GBS data during the live data analysis session they should bring:

A suitable laptop with at least an hour's battery life, loaded with:

• R v3.2.1 from http://cran.csiro.au/ • Rstudio from https://www.rstudio.com/products/rstudio/download/ • Installed R libraries: 'compiler', 'ape', 'RColorBrewer', 'NCBI2R'

This is entirely optional and is NOT necessary to take to part in the session.

22 2015 AGTA Conference

discovery and genome-directed stratified SUNDAY clinical trials for a variety of recalcitrant 11 October 2015 cancers. ABSTRACT It is now well established that a cancer’s mutational burden drives tumour formation, ORATION influences disease progression and can dictate sensitivity to chemotherapy. Over the last 6 years, we have pioneered large-scale cancer genome & transcriptome sequencing and methylome profiling of approximately Chaired by Dr Ruby Lin 700 pancreatic, Ovarian and Oesophageal patients, primarily as part of the International 1700-1800 Cancer Genome Consortium (ICGC). These data have been used to address three STUDYING CANCER AT SINGLE core questions: i) What are the mutational NUCLEOTIDE RESOLUTION mechanisms responsible for each patient’s PROFESSOR SEAN GRIMMOND tumour, ii) What are the core pathway- Wolfson Wohl Cancer Research Centre, mutations promoting each individual’s Institute for Cancer Sciences, tumour initiation, disease progression University of Glasgow, Scotland, U.K and chemo-resistance, and iii) how can personalised cancer-genome information BIOGRAPHY be used to improve patient outcome? These Over the last 20 years Prof Grimmond’s efforts have provided the foundations for a research has focused on discovering the new program focused on genome-directed underlying genetics controlling key biological clinical trials in advanced pancreatic cancer processes (pluripotency and organogenesis) patients. The strategies behind the transition and pathological states (cancer) through from cohort-based to personalised genome genomic & transcriptomic approaches. sequencing and its clinical application will be He has a broad range of scientific discussed. achievements which include:- i) mapping out the complexity and dynamics of the mammalian transcriptome, ii) studying the role of RNA mediated control in cancer and development, iii) pioneering of transcriptome and cancer genome sequencing technology, iv) resolving root causes and driving events behind Pancreatic, Ovarian and Oesophageal Cancer and v) leading Australia’s International Cancer Genome Consortium program. In 2013, Prof Grimmond moved to the UK and recently became a founding co-chair of the Scottish Genomes Partnership; an ambitious patient sequencing program to drive diagnosis of rare genetic disease, cancer genome

23 2015 AGTA Conference

and cell differentiation, primarily through MONDAY single-cell genomics. The lab will operate 12 October 2015 at the interface between genomics and experimental cell biology to answer how cells make fate decisions.

ABSTRACT Single-cell trajectory analysis is a powerful SESSION 1 approach for studying gene regulatory changes during cell differentiation and other dynamic processes. Recently, we showed that individual cells can be ordered according to progress through differentiation TRANSCRIPTOMICS AND THE by analyzing their transcriptomes with REGULATION OF THE GENOME unsupervised algorithms. Previous studies by our group and others have been limited Chaired by Dr Phillippa Taberlay and to linear trajectories tracking unipotent Dr Timothy Mercer progenitor cells. Such cellular trajectories have only one outcome. However, during 0845-0930 development, cells make fate decisions that lead to one of several mutually exclusive DIFFERENTIAL ANALYSIS OF states in the adult. How to reconstruct and BIFURCATING SINGLE-CELL GENE analyze single-cell trajectories that include EXPRESSION TRAJECTORIES and span fate decisions is an open problem.

ASSISTANT PROFESSOR COLE Here, we describe an approach for TRAPNELL reconstructing single-cell trajectories University of Washington, U.S.A that include bifurcations corresponding to cell fate decisions. We then describe BIOGRAPHY statistical methods for identifying genes Dr. Trapnell studies stem cells and that are differentially expressed between differentiation, primarily using high trajectory outcomes. We illustrate the throughput transcriptome sequencing. He power of this technique by analyzing earned his Ph.D. in Computer Science from differentiating bronchoalveolar progenitor the University of Maryland, College Park, cells undergoing specification into type I and where he was jointly advised by Steven type II pneumocytes. This analysis reveals Salzberg and Lior Pachter. As a postdoc hundreds of genes with lineage-dependent in John Rinn’s lab at Harvard’s Stem Cell expression. Our approach, which encodes and Regenerative Biology department, a topological description of the trajectory as he pioneered methods for analyzing continuous predictors in a generalized linear differentiation with single cell transcriptome model, can distinguish, for example, genes sequencing. He is the principal developer that become lineage-dependent proximal of several widely used open-source to the fate specification from those that are software tools for analyzing high-throughput restricted to a lineage later in differentiation. sequencing experiments. At the University We conclude with an analysis of bifurcations of Washington, his lab will focus on finding in settings other than development to argue genes that govern stem cell maintenance that single cell trajectory analysis can help

24 2015 AGTA Conference pinpoint the genes that drive a process from 0955-1015 those more downstream. ALTERATIONS IN THREE-DIMENSIONAL 0930-0955 ORGANISATION OCCUR COINCIDENT WITH LONG-RANGE EPIGENETIC INSIGHTS INTO MAMMALIAN CELL TYPES DEREGULATION OF THE CANCER GENOME PROFESSOR ALISTAIR FORREST Harry Perkins Institute of Medical Research, DR PHILLIPPA TABERLAY WA, Australia Garvan Institute of Medical Research, NSW, Australia BIOGRAPHY Alistair Forrest is a research professor at the BIOGRAPHY Harry Perkins Institute of Medical Research, Dr Taberlay earned her PhD in Medicine UWA. He is an expert in transcriptomics from the University of Tasmania under the and is most well-known for his work within supervision of Dr Adele Holloway. She then the Functional annotation of mammalian joined the lab of Professor Peter Jones as genomes projects (FANTOM2-5). He is the a postdoc at the Norris Comprehensive scientific coordinator of the FANTOM5 project. Cancer Center at the University of Southern This work, which profiled a large collection of California, Los Angeles where she primary cell types, cell lines and tissues using discovered that distal regulatory elements cap analysis of gene expression (CAGE), has were important for epigenetic plasticity by generated landmark maps of promoters and demonstrating that epigenetic signatures transcribed enhancers in human and mouse of enhancers determined whether a genomes. Prof Forrest is a multidisciplinary cell could be trans-differentiated by (for researcher with a BSc in Biotechnology from example, from skin to muscle). She returned Murdoch University, a Masters in information to Australia and joined the Epigenetics technology from Queensland University of Research Laborotory of Professor Susan Technology and a PhD from the Institute Clark to continue investigating epigenomic for Molecular Bioscience, University of remodelling and enhancer dynamics in Queensland. In 2007, he joined RIKEN in prostate and breast cancers. Her group Yokohama Japan on a CJ Martin Fellowship, is interested in long-range interactions where he was later promoted to team leader and understanding the role of the three- of a large bioinformatics group. He has dimensional genome in normal and recently returned to Western Australia to cancer biology in the context of epigenetic head the newly formed Systems Biology remodelling and cell reprogramming. and Genomics group. His current research concentrates on mammalian systems biology ABSTRACT with a focus on transcriptional regulatory A three-dimensional chromatin state networks (TRNs) and cell-cell communication underpins the structural and functional networks (CCCNs). basis of the genome by bringing regulatory elements and genes into close spatial proximity to ensure proper, cell-type specific gene expression profiles. Here, we perform HiC chromosome conformation to investigate how the three-dimensional organization of the prostate cancer genome is disrupted in the context of

25 2015 AGTA Conference epigenetic remodelling and atypical gene measured genome-wide, by a custom RNA- expression programs. We find that the seq approach she has developed called general feature of a chromatin interactome, PAT-seq. Traude will discuss unpublished data namely segmentation into megabase-sized that shows how changing poly(A)-tail length- topologically associated domains remain a distributions can uncover previously hidden conserved feature of cancer cell genomes, mRNA fate decisions. however these domains are generally smaller. Domain boundaries are enriched ABSTRACT for insulator binding protein CTCF and Changes to RNA metabolism are more and histone mark H3K4me3 in both normal and more implicated in in human health and cancer cells, indicating that these factors disease. An example is the recent discovery are involved in the establishment and of a wholesale switch to shorter 3’UTRs conservation of topologically associated in many cancers. This results in mRNA domains. Importantly, we also identify novel with reduced scope for post-transcriptional prostate cancer-specific interactions that regulation, such as microRNA-mediated are enriched for various regulatory elements repression. In the case of cancer, this (enhancers, promoters and insulators). means a worse prognosis for patients as These differential interactions are commonly oncogenes become deregulated (Xia et retained within the confines of topological al, 2014). In addition to the changes to the domains, are associated with differentially site of polyadenylation; a second dynamic expressed genes and overlap long-range feature of 3’UTRs is the length-control of epigenetically activated or silenced regions the Poly(A)-tail. Nascent mRNA undergo in prostate cancer. Finally, we present a a poly(A) length-check as part of quality novel visualisation tool to explore the HiC control prior to nuclear export. However, interaction in relation to the transcription and once in the cytoplasm, long poly(A)-tails epigenetic changes between normal and can be rapidly trimmed to dial down protein prostate cancer cells. translation, or to initiate decay. Our research capitalises on a new RNA-seq approach 1015-1030 developed in our lab called Poly(A)-Test- seq (PAT-seq)(Harrison et al, 2015). The 3’UTR DYNAMICS PREDICT mRNA FATE PAT-seq approach accurately records i) mRNA abundance, ii) polyadenylation site

DR TRAUDE BEILHARZ choice and iii) poly(A)-length distribution Monash University, VIC in eukaryotic transcriptomes in a single BIOGRAPHY statistically robust assay. We have applied Traude Beilharz heads the RNA-systems this approach to better understand how biology laboratory at Monash University where encrypted information in the 3’UTR of she and her team study the mechanisms of mRNA directs when, where and how- global RNA-based regulatory control. Her often mRNA is translated. We will present research is focussed on post-transcriptional unpublished PAT-seq data addressing of gene expression with a special interest questions fundamental to understanding in the 3’ UTR dynamics of eukaryotic the control of post-transcriptional gene transcriptomes. That is, how regulated expression. For example, we will show how changes to polyadenylation-site (alternative changes to poly(A)-tail length-distribution polyadenylation) and changes to steady-state can identify the substrates of specific RNA poly(A)-tail length distribution can predict binding proteins and regulatory enzymes. mRNA fate. These changes are conveniently Specifically, showing for the first time that

26 2015 AGTA Conference the binding of the pumilio domain protein combining RNA-seq analysis with LTP Puf3 to its target mRNA specifically ‘marks’ induction in the dentate gyrus of live rats, these with changes to their adenylation- we provide the first global transcriptomic state in budding yeast. Moreover we show analysis of synaptic plasticity in the adult that such changes to poly(A)-tail length- brain. Expression profiles of mRNAs and distribution can also identify the substrates long noncoding RNAs (lncRNAs) were of regulatory cytoplasmic polyadenylation; a obtained at 30 minutes, 2 hours and 5 process commonly utilised in the brain and hours after high-frequency stimulation of the germ line to activate translation. Importantly, perforant pathway. The temporal analysis the measure of these changes in 3’UTR revealed dynamic expression profiles of dynamics are often more sensitive in their lncRNAs consistent with involvement in read-out of mRNA fate than traditional LTP. In light of observations suggesting a antibody-based approaches. role for retrotransposons in brain function, we examined the expression of various 1030-1045 classes of repeat elements. Our analysis identifies dynamic regulation of LINE1 and DYNAMIC EXPRESSION OF LONG SINE retrotransposons, and tRNA. These NONCODING RNAs AND REPEAT experiments reveal a hitherto unknown ELEMENTS IN SYNAPTIC PLASTICITY complexity of gene expression in long-term MR JESPER MAAG synaptic plasticity involving the dynamic Garvan Institute of Medical Research, NSW regulation of lncRNAs and repeat elements. Our analyses provide a foundation for BIOGRAPHY understanding the molecular underpinnings Jesper Maag is a third year PhD-student in of synaptic plasticity in both the healthy the Genome Informatics lab under A/Prof brain and in neurodegenerative and Marcel Dinger at the Garvan Institute of neuropsychiatric disorders. The dynamic Medical Research. Jesper completed a master expression of retrotranspons points towards degree in pharmaceutical sciences at Uppsala a mechanism for genetic somatic mosaicism University, Sweden. He wrote his master’s observed in the brain and supports the idea thesis in the pharmacology department at the that dynamic reprogramming of the genome University of Adelaide, where he explored the may be an important mechanism underlying effect of cannabinoids on Alzheimer’s disease. memory formation. Jesper continued his academic career with a transition to a PhD in Genomics in Sydney.

ABSTRACT Long-term potentiation (LTP) of synaptic transmission is recognized as a cellular mechanism for learning and memory storage. Although de novo gene transcription is known to be required in the formation of stable LTP, the molecular mechanisms underlying synaptic plasticity remain elusive. Non-coding RNAs have emerged as major regulatory molecules that are abundantly and specifically expressed in the mammalian brain. By

27 2015 AGTA Conference

Abstract The genomes of our nearest extinct human relatives, the Neandertals and Denisovans, SESSION 2 offer a unique opportunity to identify genetic changes that have come to fixation or reached high frequency in modern humans since their divergence from Neandertals and GENOME EVOLUTION AND Denisovans, as well as regions of modern BIOINFORMATICS human genomes that have been impacted by admixture with archaic humans. Some of these changes may have important Chaired by Dr Natalie Thorne and Dr Joseph functional effects in modern humans. Powell We have sequenced to high-coverage 1115-1200 the genomes of two archaic hominin groups: Neandertals [1] who lived in Europe WHAT WE HAVE LEARNED FROM and Western Asia until approximately 30 SEQUENCING ARCHAIC HUMAN 000 years ago, and Denisovans [2], a group GENOMES related to Neandertals that was recently described based on the genome sequence DR JANET KELSO generated from a bone found in Southern Max Planck Institute for Evolutionary Siberia. Using these genome sequences we Anthropology, Germany have identified gene flow between archaic BIOGRAPHY and modern humans. About 2.0% of the Janet Kelso is the Bioinformatics research genomes of present-day non-Africans derive group leader at Max-Planck Institute for from Neandertals, while about 4.8% of the Evolutionary Anthropology in Leipzig. genomes of present-day Oceanians derive from Denisovans. Her research interests include analysis of ancient genomes, primate comparative A map of the Neandertal haplotypes in genomics and gene expression. Her group present-day non-Africans has allowed uses computational approaches to gain us to characterize some of the functional insights into genome evolution and have a impacts of Neandertal ancestry on modern special interest in the development of novel humans [4], while the lengths of these software for processing and analysis of high- introgressed regions in an early modern throughput sequence data. human have provided information about the date of admixture. Further we have Janet received her PhD from the South identified sequence differences that have African National Bioinformatics Institute come to fixation or reached high frequency under the supervision of Winston Hide. She in modern humans since their divergence is author of more than 50 peer-reviewed from Neandertals and Denisovans, some of scientific publications, and is currently the joint which may have important functional effects Executive Editor of the journal, Bioinformatics in modern humans. and an Associate Editor of the journal, Database.

28 2015 AGTA Conference

1200-1225 sufficient weight of evidence to be able to be declared as new epilepsy genes. USING GENE CO-EXPRESSION We have made use of several large-scale NETWORKS TO DISCOVER EPILEPSY gene expression datasets to show that GENES (i) known epilepsy genes co-express in a limited number of sub networks, and (ii) ASSOCIATE PROFESSOR MELANIE these networks can be used to prioritise BAHLO which genes to target for further testing. Walter + Eliza Hall Institute of Medical Our approach has been very successful Research, VIC for epileptic encephalopathies (Oliver et al BIOGRAPHY 2014, PLoS ONE). Performance evaluation Associate Professor Melanie Bahlo is the of the prioritization approach applied to co-Division Head of the Population Health 179 candidate genes has seen six of these and Immunity Division at the Walter and genes now validated as true epileptic Eliza Hall Institute of Medical Research. encephalopathy since 2014, with all but one She graduated in 1998 with a PhD in prioritized by us. We have now expanded population genetics from Monash University, this approach to encompass all available Melbourne. She currently holds a National large brain expression microarray datasets, Health and Medical Research Council further increasing the power of this method, Senior Research Fellowship. In 2009 she which also works in other neurogenetics was awarded the Moran Medal by the disorders such as brain malformation Australian Academy of Science and in disorders. 2015 the Genetics Society of Australasia’s Ross Crozier medal. Her research areas 1225-1245 cover statistical genetics, bioinformatics DEEP GENE EXPLORATION OF HUMAN and population genetics with a focus on CHROMOSOME 21 neurological disorders and infectious disease. Her statistical analyses have led to DR TIMOTHY MERCER the identification of novel genes for disorders Garvan Institute of Medical Research, such as epilepsy and deafness. She has NSW, Australia also developed new methods and software for the analysis of genetic data. BIOGRAPHY Dr Timothy Mercer undertakes research Abstract into RNA biology (noncoding RNAs, gene Epilepsy is a disease which forty years organization, expression and splicing), ago wasn’t even thought to have a genetic genome informatics and bioinformatic and basis. It can be further subcategorized sequencing innovations. He currently leads phenotypically, with some subphenotypes a research group at the Garvan Institute of showing strong Mendelian segregation. Medical Research, Sydney. To date more than one hundred epilepsy genes have been discovered, leading to Abstract the recognition that epilepsy is genetically The human genome encodes an unknown highly heterogeneous. Large NGS studies diversity of protein-coding and noncoding have been highly successful, in particular RNAs. The vast diversity and wide with some clinical subphenotypes, such as dynamic range of gene expression limits epileptic encephalopathies. NGS cohort the capacity of RNA sequencing to resolve studies yield many hits that do not have the transcriptome and, as a result, current

29 2015 AGTA Conference reference catalogs provide an incomplete Brian is involved in a number of collaborative profile of human gene content. Here, projects world-wide using existing and we target RNA sequencing to human emerging transcriptomic technologies chromosome 21 and syntenic regions of to answer key questions in normal the mouse genome to obtain saturating development and disease processes. coverage over a cross-section of the mammalian transcriptome. This reveals He currently holds a Cancer Institute of the full size, structure and composition of NSW early career fellowship and a conjoint coding and noncoding RNA populations and lecturer position at UNSW. enables interrogation of their expression, Brian has published eight papers in the splicing and evolution at unprecedented last five years including three first and two resolution. We show that noncoding exons, second author publications. He also holds in contrast to coding exons, are universally an international patent for a novel method alternatively spliced to generate a near- of using epigenetic marks for diagnosing inexhaustible diversity of isoforms. This cancer. In 2012, he received his PhD for isoform diversity is regulated by local cis- work on the whole genome identification elements that are sufficient to recapitulate and characterisation of novel epigenetically human alternative splicing profiles in a silenced genes in ovarian cancer. mouse cell. This is despite the noncoding RNA populations of human and mouse Abstract having largely diverged, with orthologs rare Despite increasing recognition of the role between the two species. Together, these that long noncoding RNAs (lncRNAs) play findings demonstrate that the coding and in the regulation of gene expression, little is noncoding RNAs are fundamentally distinct known about their dynamic changes, which in their organization and evolution and in turn limits the understanding of how and forecasts the full breadth and diversity of where these molecules act within regulatory genes encoded in the human genome. networks. Investigations of transcriptional responses during biological processes 1245-1300 commonly use transcriptome profiling by microarray or RNA-Seq over 24-hourly time- MAPPING STEM CELL BIOLOGY courses. However, to date such experiments BY HIGH RESOLUTION TEMPORAL are typically designed around principles to TRANSCRIPTOMICS examine protein-coding transcripts, limiting DR BRIAN GLOSS their potential to explore the transcriptional Garvan Institute of Medical Research, dynamics of lncRNAs. In addition to their NSW, Australia relatively poor annotation and difficulty of detection due to low expression levels BIOGRAPHY compared to protein-coding transcripts, Brian Gloss has been a postdoctoral lncRNAs are turned over more rapidly than researcher in the Genome Informatics mRNAs. Therefore, we hypothesized that research lab at the Garvan Institute under A/ existing transcriptome profiling experiments Prof Marcel Dinger since 2013. His research were inadequately powered in terms of both interests revolve around unravelling temporal resolution and sequencing depth transcriptional complexity in health and to gauge the majority of their dynamics and disease. that rapidly turned over regulatory transcripts may have evaded detection entirely. To test this hypothesis, we evaluated coding

30 2015 AGTA Conference and noncoding gene expression dynamics in differentiating mouse embryonic stem cells at unprecedented temporal resolution (6-hourly) and sequencing depth to examine SESSION 3 the effects of this increased experimental power on the characterization of the molecular processes underlying stem cell differentiation. We show that many key DEVELOPMENTAL GENOMICS AND transcriptional changes, including regulatory THE ORIGINS OF DISEASE network interactions, gene expression changes in coding and noncoding genes, periodically expressed genes, and Chaired by Professor Andrew Perkins and alternative splicing events, cannot be derived Dr Kate Howell from existing developmental time-courses. Moreover, we identify hundreds of novel 1530-1555 transiently expressed lncRNAs that are only detectable by high frequency sampling. As HIGHLY CONSERVED EPIGENOME these findings have important implications REMODELLING DURING THE upon the investigation of regulatory VERTEBRATE PHYLOTYPIC PERIOD dynamics and networks of any biological PROFESSOR RYAN LISTER process, we present experimental design The University of Western Australia, WA considerations and methodologies that can be utilized in future studies. BIOGRAPHY Ryan Lister leads a research group exploring the epigenome at the University of Western Australia and the Harry Perkins Institute of Medical Research. After receiving his PhD from UWA, Ryan undertook postdoctoral studies at The Salk Institute for Biological Studies from 2006. There he developed methodologies for utilizing new high- throughput DNA sequencing technologies to map the epigenome, the molecular code superimposed upon the genome that plays crucial roles in regulating the information contained in the underlying DNA sequence. His research has yielded new insights into the composition and function of the epigenome in a variety of systems, including plants, the brain, and stem cells. Having returned to UWA in 2012, Ryan’s laboratory is focused upon understanding how these complex epigenomic patterns are established and altered, how they affect the readout of underlying genetic information, their role in brain development and function, and developing molecular tools to precisely edit the epigenome.

31 2015 AGTA Conference

Abstract Genetic Clinics at The Children’s Hospital The vertebrate body plan and organs are at Westmead, and the Westmead Adult shaped during a highly conserved embryonic Hospital, Sydney. Her group’s work has phase called the phylotypic stage, however been integral in establishment of genomic the mechanisms that guide the epigenome approaches for genetic diagnosis of eye through this transition and their evolutionary diseases in Australia, and development of conservation remain elusive. Here we animal and cell-based systems for functional report widespread DNA demethylation understanding of novel disease genes and of thousands of enhancers during the variants, and pathways to treatment. phylotypic period in zebrafish, Xenopus and mouse. These dynamic enhancers are Abstract linked to essential developmental genes Many of the genetic factors and their that display coordinated transcriptional functions contributing to human disease and epigenomic changes in the diverse are not yet known. Genomic strategies vertebrates during embryogenesis. provide an unprecedented opportunity Phylotypic stage-specific binding of Tet for identification of novel factors and proteins to (hydroxy)methylated DNA, and mechanisms causing human disease. We enrichment of hydroxymethylcytosine on use interrogation of genomic structural these enhancers, implicated active DNA rearrangements and whole genome demethylation in this process. Finally, sequencing for single pathogenic variant disruption of Tet activity demonstrated detection, to identify novel regulatory that this highly coordinated epigenome mechanisms and candidate disease reconfiguration event is dependent upon genes. We use a functional genomics the Tet proteins. Overall, this reveals a pipeline, including CRISPR/Cas9 genome novel regulatory module required during engineering in cell-based and animal model the most conserved phase of vertebrate studies, to determine pathogenicity of embryogenesis and uncovers an ancient the identified genomic variants and novel developmental role for the Tet dioxygenases. disease genes. Our studies reveal new factors critical in cell polarity and adhesion, 1555-1615 epithelial to mesenchymal transition, cell division, ciliogenesis and regulation of GENOME-DRIVEN INSIGHTS TO DISEASE Wnt signalling. Correlation with patient MECHANISMS phenotypes facilitates prioritization of functional studies, aimed at development of ASSOCIATE PROFESSOR ROBYN treatments for the conditions. JAMIESON Children’s Medical Research Institute, NSW

BIORGRAPHY Associate Professor Robyn Jamieson is the Genomics Research Lead of the Western Sydney Genetics Program at The Children’s Hospital at Westmead, Sydney, and heads the Eye Genetics Research Group at the Children’s Medical Research Institute, The Children’s Hospital at Westmead and Save Sight Institute, University of Sydney. She is also the Lead Clinical Geneticist at Eye

32 2015 AGTA Conference

1615-1635 investigate the genetic networks regulating any specific biological process. A FUNCTIONAL GENOMICS APPROACH TO PREDICT NOVEL We have a long-standing interest in GENETIC DETERMINANTS FOR HEART unravelling the exact genetic components DEVELOPMENT AND DISEASE that build a healthy heart. In a bid to systematically investigate the players DR MIRANA RAMIALISON involved in cardiogenesis, we took Australian Regenerative Medicine Institute, advantage of our unique bioinformatics Monash University, VIC expertise to mine the publicly available cardiac-specific datasets recently released BIOGRAPHY by the consortia. We discovered regulatory Dr. Ramialison is head of the Systems patterns that are specific to genes controlling Developmental Biology Laboratory at the heart development. We used these patterns Australian Regenerative Medicine Institute to predict novel genes having an essential in Melbourne. She is an NHMRC/NHF function in cardiogenesis and are currently Career Development Fellow and leads a validating the function of these genes using multi-disciplinary team of bioinformaticians the mouse and zebrafish model systems. and molecular biologists, to study heart Given that 80% of Congenital Heart Disease development, evolution and disease. (CHD) cases are of unknown genetic cause, She takes a systems biology approach to this study will expand our knowledge on uncover the gene regulatory networks that genes responsible for CHD. control gene expression during cardiac development, and identify abnormal 1635-1650 interactions that cause congenital diseases. Prior to joining ARMI in February 2014, REPLI-G SINGLE CELL SEQUENCING: Dr Ramialison was an EMBO and HFSP DRIVING THE NEXT FRONTIER OF post-doctoral Fellow at the Victor Chang GENOMIC AND TRANSCRIPTOMIC Cardiac Research Institute in Sydney. She ANALYSIS received her Engineering degree from the University of Luminy (France) and PhD at MR COLIN BARON the European Molecular Biology Laboratory QIAGEN LIFE SCIENCES, USA (Germany).

ABSTRACT In 2001, the release of the first human genome draft sequence provided a breakthrough in the knowledge of our genetic repertoire, albeit the function of BIOGRAPHY the newly discovered genes remained Colin Baron currently serves as the Sr. largely unknown. In 2012, international Director of Product Management at Qiagen consortia (ENCODE, FANTOM) released Life Sciences and the Head of NGS & Single the first functional map of the mouse and Cell Technologies. Prior to joining Qiagen, human genomes, thereby providing an Colin worked for 6 years at Illumina where exhaustive picture of gene expression and he led product management teams covering gene regulation in a temporal and tissue- Ilumina’s core sequencing chemistries and specific manner. These multi-million dollar was responsible for the reagent launches public efforts enable us to systematically for all major platform releases from the early

33 2015 AGTA Conference days of the Genome Analyzer to the launch sequencing library preparation can be of the HiSeq X 10 and NextSeq systems. integrated into typical High Throughput and Benchtop NGS systems, and affords highly Colin holds degrees in Microbiology and uniform sequence coverage and high library Economics and a Master of Business complexity with minimal sequence bias. Administration from the University of California, Davis and worked for 8 years at Highlighting research from our customers, the UC Davis Cancer Center where he led we will demonstrate how high accuracy, research teams studying Copy Number and single cell sequencing can cut through the gene dosage effects in children with Autism. biological noise that occurs when analyzing heterogeneous, bulk samples. With lower ABSTRACT sequencing depth requirements, researchers Single-cell genomic analysis enables can scale the size and statistical power of researchers to gain novel insights across their studies. a wide array of applications, including developmental biology, tumor heterogeneity 1650-1705 and disease pathogenesis and progression. Typically, conducting single-cell genomic UNCOVERING HIDDEN GENES IN analysis using next-generation sequencing INTERGENIC GWAS REGIONS (NGS) methods is challenging because the amount of genomic DNA present in a Dr Nenad Bartonicek Garvan Institute of Medical Research, NSW, single cell is very limited. PCR-based whole Australia genome amplification methods tend to have high error rates, low coverage uniformity, BIOGRAPHY extensive allelic drop-outs and limited Nenad Bartonicek is a postdoctoral amplification yields. researcher in the group of Marcel Dinger at Garvan Institute of Medical Research since Qiagen has designed a portfolio of Single 2013. He received his PhD from University Cell products for genomics applications, of Cambridge, UK while working at the including NGS, using our unique Multiple European Bioinformatics Institute. Displacement Amplification (MDA) approach that eliminates PCR-related bias and ABSTRACT provides a high accuracy view into the The majority of SNPs linked with diseases genome and transcriptome diversity between through genome-wide association studies cells. (GWAS) reside in intergenic regions. These “gene deserts” can result from the blind spot We describe a new, streamlined workflow of current sequencing methods for lowly for single-cell NGS library construction that abundant and tissue specific transcripts. In uses Multiple Displacement Amplification order to address the hidden transcription in (MDA) to amplify the whole genome with linkage disequilibrium blocks around GWAS high uniformity and fidelity, and library SNPs, we employed the recently developed construction with high adaptor ligation RNA capture sequencing (CaptureSeq) efficiency. The entire procedure generates on 21 human tissues. CaptureSeq was high-quality sequencing libraries without designed to enrich for 560 intergenic GWAS PCR amplification, thereby eliminating regions, covering 2.3 percent of the genome. PCR-related bias and errors and reducing After transcript assembly, quantification handling steps. Our data demonstrate and filtering, we identified ~1000 novel, that this PCR-free method for single-cell high-confidence (multi-exonic, FPKM>10)

34 2015 AGTA Conference

genes, with 80% regions containing at ABSTRACT least one high-confidence transcript. The Functionally annotating genomic variants genes contained major hallmarks of active is a crucial step in elucidating the transcription, such as enrichment for typical biological processes that manifest in epigenetic marks and CAGE tags, and were certain phenotypes, such as disease. A similar to other ncRNAs in composition of commonly employed annotation is whether isoforms, exons and splice junctions. About a variant results in a non-synonymous 50% of transcription loci were previously amino acid substitution and whether they observed in EST databases, although mostly are predicted to be deleterious. However, with only a fraction of sequence overlap while this is likely conclusive in cases of (<5%). In total, 5500 high-confidence protein truncations, it may not offer insights transcription events were observed across into variants causing structural changes. all tissues, with 232 genes showing tissue This is particularly the case for instances specificity. Finally, we used information where seemingly deleterious variants co- from 13 melanoma samples in combination mutate resulting in matching changes of the with TCGA studies to identify a number of interacting amino acids in the 3D structure. previously unreported melanoma associated The first step towards elucidating such transcripts. complex interactions is to visualize the variant locations within a protein structure 1705-1720 to provide more insight as to potential consequences. VISUALIZATION OF VARIANTS USING AQUARIA: NON-SYNONYMOUS SNPs IN We therefore introduce Visualization of THE CONTEXT OF PROTEIN STRUCTURE Variants using Aquaria (VVA), a web- based tool to assist researchers with this

DR NEIL SAUNDERS task. While previous approaches focus Digital Productivity Flagship, CSIRO, NSW on visualizing preprocessed SNP data [1- BIOGRAPHY 4], VVA allows the user to provide a list of Neil did his Ph.D. in Biochemistry at the patient-specific variants (such as a VCF file) University of Oxford and his first postdoc in containing their chromosomal coordinates the Department of Molecular Cell Physiology as well as a protein ID of interest. The at the Free University of Amsterdam. He Ensembl REST API is then employed to moved to the University of New South Wales label which of the supplied variants result in in 2000 and commenced a second career non-synonymous amino acid mutation for in bioinformatics, working on the genomes the protein of interest. Genomic (nucleotide) of Archaea isolated from Antarctica. In 2006 coordinates are converted to amino acid he moved to the University of Queensland sequence positions and the mutation data to work on the computational prediction are uploaded to Aquaria [5], so as the variant of protein-protein interactions. He has location within the corresponding protein been with CSIRO since 2009, where he structure can be visualized. is currently a statistical bioinformatician in We will outline the development of the the Transformational Bioinformatics Team, variant visualization system and present part of the eHealth Research Program, specific examples highlighting the value of and works on the analysis of large “omics” visualizing variants in the context of the 3D datasets for human health. protein structure.

35 2015 AGTA Conference

genome sequenced breast cancers could TUESDAY reveal mutation signatures, imprints left by 13 October 2015 mutagenic processes that have occurred through cancer development. In particular, she identified a novel phenomenon of localised hypermutation termed “kataegis”. SESSION 4 In 2013, Serena was awarded a Wellcome Trust Intermediate Clinical Fellowship to pursue biological understanding of the signatures identified during her research training. She joined the Sanger CANCER GENOMICS AND Institute faculty team in 2014 and leads the Signatures of Mutagenesis group. PRECISION MEDICINE Serena continues to hunt for mutation signatures in large cancer datasets using computational approaches. She explores Chaired by Professor David Thomas and these signatures biologically through cell- Dr Nic Waddell based model systems. Serena runs a clinical 0830-0910 project, Insignia (www.mutationsignatures. org), recruiting patients with DNA repair/ ADVANCES IN THE UNDERSTANDING OF replication defects, aging syndromes and MUTATIONAL SIGNATURES IN HUMAN neurodegeneration, and people who have SOMATIC CELLS been exposed to environmental/occupational mutagens to gain biological insights into DR SERENA NIK-ZAINAL mutational phenomena in these patients. Wellcome Trust Sanger Institute, Cambridge Abstract BIOGRAPHY Somatic mutations in human cancers are Serena is a Career Development Fellow generated by multiple mutational processes (CDF) Group Leader in the Cancer Genome operating at various times in the cellular Project and an Honorary Consultant in lineage between the fertilized egg and the Clinical Genetics at Addenbrooke’s Hospital cancer cell, each composed of specific DNA in Cambridge. She is pursuing biological damage and repair components and leaving understanding of the mutational signatures its own characteristic imprint, or mutational that have been identified in primary human signature, on the genome1,2. Exploiting cancers. the scale offered by modern sequencing Serena qualified in medicine from the technologies, complete catalogues of somatic mutation can be comprehensively University of Cambridge in 2001, trained as 3 a physician and subsequently specialised explored and mathematical methods in Clinical Genetics. Serena undertook a have been employed to identify mutational signatures in human cancer (http://cancer. PhD at the Wellcome Trust Sanger Institute 1-3 (WTSI) in 2009 with Mike Stratton exploring sanger.ac.uk/cosmic/signatures) . breast cancer using next-generation We now continue to explore the breadth sequencing (NGS) technology. She and depth of mutation signatures in somatic demonstrated how detailed downstream cells. In particular, I will present results analyses of all mutations present in whole- from an analysis of 560 breast cancers,

36 2015 AGTA Conference

the largest collection of whole genome Abstract sequenced cancers of a single cancer type The Druley lab is focused on the to date. This powerful dataset reveals novel translational integration of computational and insights into how DNA repair, replication, molecular biology to study the connection transcription and chromatin organization between human development and pediatric influences breast mutagenesis. I will also cancer. Multiple studies suggest that explain our plans to further the biological most pediatric cancers lack the number of understanding of mutational signatures in somatic mutations necessary for a single somatic cells4. cell to transform to malignancy, suggesting other factors are required. In addition, 0910-0945 there is growing recognition that children born with congenital anomalies have a DEVELOPMENTAL ANOMALIES AND higher incidence of childhood cancer and PEDIATRIC CANCER: A GROWING as many as one-third of pediatric cancer CONNECTION patients are predisposed to cancer. These DR TODD DRULEY observations have led us to develop Washington University School of Medicine, genomic assays that allow us to explore U.S.A profiles of germline variation, which disrupt normal human developmental mechanisms BIOGRAPHY and result in uncontrolled proliferation. Our Dr Druley is a faculty member in the Division initial model to test this hypothesis is infant of Pediatric Hematology and Oncology. His leukemia (IL), which is the deadliest of all research is based in the hypothesis that pediatric leukemias with <50% survival. much of pediatric disease presentation IL arises in utero and presents equally as and treatment is heavily influenced by an ALL or AML with ~75% of cases harboring individual’s unique combination of rare, a rearrangement in MLL. Yet these inherited DNA variations. Dr. Druley’s translocations alone are unable to generate laboratory focuses on adapting the latest leukemia in animal models when expressed genomic technologies toward characterizing at physiologic levels. We find that IL patients these germline variants on a population- harbor significant germline variation in based scale as well as examining how these genes comprising the COMPASS-like mutations interact within an individual to complexes, which regulate gene expression influence a variety of diseases, particularly via histone modification during mesoderm pediatric and infant leukemia. The long- and hematopoietic development in model term goals are to better understand the systems. Using patient-specific iPS lines, pathophysiology of pediatric cancer as well we can model hematopoietic differentiation as to identify all of the pertinent germline and characterize genomic targets of histone variants within a pediatric cancer patient modification, chromatin accessibility and prior to starting therapy in order to provide subsequent gene expression leading to a genetically customized treatment plan infant leukemogenesis. Using cell free DNA designed to maximize anti-cancer activity in maternal circulation, these genetic profiles while minimizing toxicity and morbidity. would improve prenatal risk prediction as well as postnatal surveillance and therapeutic selection.

37 2015 AGTA Conference

0945-1015 mechanism; to guide the interpretation of coding variants in genomes; and to better HIGHLY PARALLEL MEASUREMENT understand protein evolution. OF THE IMPACT OF MUTATIONS IN PROTEINS 1015-1030 ASSISTANT PROFESSOR DOUGLAS GENOMIC CATASTROPHES DRIVING FOWLER TUMORIGENESIS IN ESOPHAGEAL Department of Genome Sciences, University ADENOCARCINOMA of Washington, U.S.A DR KATIA NONES BIOGRAPHY QIMR Berghofer, QLD Dr Fowler focuses on developing and implementing new technologies to address BIOGRAPHY difficult problems in protein science and Katia is a genome biologist at QIMR genomics. He is currently an Assistant Berghofer. She works on the interpretation Professor of Genome Sciences and Affiliate of next generation sequencing and array Assistant Professor of Bioengineering at the data in cancer research. She is a member of University of Washington in Seattle. the Australian International Cancer Genome Consortium (ICGC) and her research has His lab currently works to understand how been focused on using genomic, epigenomic variation in protein coding regions of the and expression data in multi-discipinary genome relates to disease, as well as to fields with focus in cancer research. Her develop new methods for understanding current research focuses on the analysis of proteins. Dr Fowler’s did his graduate work whole-genome sequencing and methylome at The Scripps Research Institute, where in esophageal adenocarcinomas to better he discovered and characterized the first understand this disease in the hope to mammalian functional amyloid protein, suggest alternative treatment opportunities. Pmel17. During his postdoctoral work at the University of Washington, he developed Abstract high-throughput methods for analyzing the Esophageal adenocarcinoma (EAC) effect of mutations on protein function. incidence is rapidly increasing in Western countries. EAC has one of the poorest Abstract outcomes of all solid tumors, with only Deep mutational scanning is a method 14% of patients surviving 5 years. A that marries selection for protein function better understanding of the EAC genomic amongst a large library of protein variants landscape underpins efforts to improve early with high-throughput DNA sequencing detection and treatment outcomes. Recent to measure the activity of hundreds of studies using exome sequencing have thousands of variants simultaneously. reported recurrent loss of function mutations The result is a sequence-function map in EAC but oncogenic driving events have that describes the impact of all possible been under-represented. To address this single and many double mutants on we used a combination of whole genome protein function. We have shown that sequencing (WGS) (n=56) and SNP-arrays sequence-function maps have many (n=71) and found that genomic catastrophes uses. For example, we are analyzing are frequent in EAC, with approximately one them to learn about protein properties like third (33%, n=42/127) of tumors showing structure, aggregation, stability and enzyme evidence of chromothripsis. We showed

38 2015 AGTA Conference

that catastrophes may lead to oncogene of the Month award at Faculty of Medicine in amplification through chromothripsis-derived UNSW September 2014. She has published double minute chromosome formation 4 manuscripts in peer-reviewed journals and (MYC and MDM2) or breakage-fusion- presented her research at many international bridge (KRAS, MDM2 and RFC3). Our conferences. Dr Sun main research focuses finding of prominent telomere shortening on cancer biology, epigenomics and in EAC samples bearing localized genomic epigenetics, as well as novel drug discovery. catastrophes, as compared to their matched normal samples, suggests that physical Abstract stress during cytokinesis could give rise to Myc oncoproteins induce cancers by binding chromosomal shattering in these samples. to Myc-responsive element E-boxes at target Mutational signatures analysis showed gene promoters, leading to transcription a high frequency of T>G mutation at TT activation. Histone H3 lysine 4 (H3K4) sites, providing some evidence of an trimethylation at target gene promoters environmental mutagenic effect. Additionally, is a strict pre-requisite for Myc-induced 5% of EAC tumors display a higher transcriptional activation. The histone H3K4 contribution of BRCA signature that suggests presenter WDR5 plays an essential role in potential defective DNA repair in a small H3K4 trimethylation. Here we identified both fraction of tumors. These findings suggest canonical and non-canonical E-boxes at the that genomic catastrophes play a significant WDR5 gene core promoter and showed that role in the malignant transformation of N-Myc up-regulated WDR5 gene expression EAC in at least a third of cases, and reveal in neuroblastoma cells. Affymetrix microarray less common mutagenic factors through studies demonstrated that WDR5 target signature analysis. genes included those with E-Boxes at promoters, such as MDM2. GSEA analysis 1030-1045 showed that knockdown WDR5 preferentially down-regulated the expression of genes THE HISTONE H3 LYSINE 4 PRESENTER with E-boxes at promoters, and ChIP-Seq WDR5 IS A BIOMARKER IN N-MYC- data revealed that repression of WDR5 INDUCED TRANSCRIPTIONAL reduced H3K4 trimethylation at Myc-binding ACTIVATION AND NEUROBLASTOMA gene promoters. Protein co-IP and ChIP PROGRESSION assay demonstrated that WDR5 formed a protein complex with N-Myc, but not p53, DR YUTING SUN at the MDM2 gene promoter, leading to Children’s Cancer Institute, NSW histone H3K4 trimethylation and MDM2 BIOGRAPHY gene transcription. We have also found Dr Yuting Sun is a Research Officer and that knocking-down WDR5 considerably Conjoint Associate Lecturer in Children’s up-regulated the expression of wild type Cancer Institute at University of New South but not mutant p53 protein, leading to Wales. She received her Medicine Degree neuroblastoma cell growth inhibition and from Harbin Medical University in China and apoptosis. In addition, WDR5 was over- worked at the Department of Gynaecology in expressed in pre-cancer ganglia and the 2nd Affiliated Hospital of Harbin Medical neuroblastoma cells from MYCN transgenic University. In 2014, she completed her mice, compared with normal ganglia cells. In PhD research in Paediatric Oncology from neuroblastoma patients, high levels of WDR5 UNSW. Dr Yuting started her post doctorate expression in tumor tissues independently since August 2014 and won the Researcher predicted poor overall survival. Our

39 2015 AGTA Conference

findings therefore identify WDR5 as an involved in the implementation of these novel important biomarker for N-Myc-regulated genomics approaches in routine clinical transcriptional activation and tumorigenesis, diagnosis. and as a novel therapeutic target for N-Myc over-expressing neuroblastoma. Abstract Rapid developments in genomics technologies now allow us to sequence all genes (the exome) or even the entire genome of thousands of patients in research SESSION 5 and diagnostics. This is completely changing the way genetics studies are done, and allows us to take full advantage of the power of genomics in medicine. Exome sequencing CLINICAL GENOMICS AND THE is now a routine diagnostic test in our FUTURE OF HEALTHCARE laboratory for more than 16 different genetic disorders, and we are in the process of moving to clinical whole genome sequencing Chaired by Dr Cas Simons and using a novel high-throughput sequencing Dr Denis Bauer platform. With this, the identification of genetic variations is not a bottleneck 1115-1155 anymore, and we can now focus on the major remaining bottleneck; Interpretation FROM GENES TO GENOMES IN MEDICAL of the enormous amount of genomic RESEARCH AND PATIENT CARE variation in the context of a clinically PROFESSOR JORIS VELTMAN heterogeneous phenotype. Solving this Radboud University Medical Centre and will require a concerted clinical, biological Maastricht University Medical Centre, and bioinformatics approach, international The Netherlands agreement on phenotype ontologies, sharing of clinical and genomic data, optimization of BIOGRAPHY variant interpretation tools and the validation Professor in Translational Genomics of these using relevant biological models. and head of Genome Research division I will demonstrate the progress made in at the Department of Human Genetics, the last few years and discuss some of the Radboud University Medical Centre in remaining issues using severe intellectual Nijmegen and the Department of Clinical disability as a model disorder. Genetics, Maastricht University Medical Centre in Maastricht, The Netherlands. My research focuses on the identification and interpretation of genomic variation, with a particular interest in the role of rare de novo mutations and copy number variations in severe neurodevelopmental and psychiatric diseases. With my research group I study the genomes of patients using next generation sequencing technology and combine laboratory experiments with novel bioinformatic approaches. I am also actively

40 2015 AGTA Conference

1155-1220 whole genome sequencing (WGS) have had a remarkably rapid, and positive impact WHOLE GENOME SEQUENCING BASED upon patients. Diagnostic rates for patients CLINICAL GENOMICS with rare monogenic diseases have risen to on average 25% under WES, and 42- DR MARK COWLEY 60% using WGS in patients with intellectual Garvan Institute of Medical Research, NSW disability. As genomics is translated into routine medicine, one key challenge will be to meet the rigorous demands of operating in a clinical environment, whilst still being agile enough to keep the process up to date with

BIOGRAPHY advances in academia, at significant patient Dr Mark Cowley is an early career fellow numbers. The incredible opportunity is that with the Cancer Institute of NSW, and the if patient genotype-phenotype data can be Team Leader of Translational Genomics, at stored in suitable databases, these datasets the Kinghorn Centre for Clinical Genomics, can be mined for research, and drive within the Garvan Institute. further improvements to patient care, and His group develops tools and approaches our understanding of the impact of genetic to comprehensively characterise the variation on phenotype. human genome, using predominantly Since the introduction of the Illumina HiSeq Whole Genome Sequencing data. We aim X, the cost of WGS has plummeted, and is to identify, understand and interpret the expected to continue to fall, making WGS- functional impact of genetic variation. The based diagnostics an attractive single- group has a strong translational focus, platform, which can be validated once, and aiming to improve clinical care through repurposed to different diseases. In order to applying genomics in a clinical environment. meet the anticipated scale of WGS-based We have focussed on using WGS to help diagnostics, we have developed a number diagnose patients with inherited genetic of key platforms, including Patient Archive, disease, particularly kidney disease, and for capturing machine-readable patient more recently, on the application of WGS to phenotype from clinical notes, Sabretooth characterise patient tumours, with the goal of for local- and cloud-based genomic analysis obtaining therapeutic insights. pipelines, and Seave, a comprehensive Mark received a BSc (Bioinformatics) from variant filtration and interpretation University of Sydney in 2003, then a PhD platform. We are currently seeking clinical from University of New South Wales in accreditation, and as such have embarked 2009. He has been at the Garvan Institute upon extensive validation of the entire since late 2008, working initially in the Peter process, from patient to clinical, both Wills Bioinformatics Centre, and then from through sequencing reference materials (eg 2011 with Andrew Biankin in the Australian NA12878), and clinical samples. In one case Pancreatic Cancer Genome Initiative and the study, for patients with Autosomal Dominant International Cancer Genome Consortium. Polycystic Kidney Disease, we obtained He has been with the KCCG since 2013. much higher diagnostic yield from WGS vs WES (86% vs 50%), due to coding variants ABSTRACT missed by WES, and multi-exon deletions Targeted sequencing, whole exome detected by WGS. Here, we will present sequencing (WES), and more recently, the platforms that we have developed, the

41 2015 AGTA Conference

findings supporting clinical validation of and uniform depth of coverage from Illumina WGS, and a number of case studies of WGS HiSeq X whole genome sequencing data applied to patients with rare monogenic presents a unique opportunity to develop disorders, or cancers with no available improved methods to detect CNVs and SVs treatment options. across the entire size gamut from hundred’s of nucleotides to entire chromosomes, in a 1220-1235 clinical setting. A comprehensive assessment of SV is challenging due to their diversity STRUCTURAL VARIATION CALLING in their manifestation, including deletions, FROM HiSeq X WHOLE GENOME duplications, inversions, translocations, SEQUENCING DATA IN A CLINICAL their size differences, their possibility to SETTING superpose forming complex rearrangements, DR ANDRE MINOCHE and their prevalence in genome repeats. Garvan Institute of Medical Research, NSW, We address this challenge by integrating Australia SV calls made from discordant read pairs, split reads and read depth, and by applying BIOGRAPHY multiple independent algorithms to detect Andre completed his Master of Science large and small SVs. Additional post-filtering in Biotechnology at the École supérieure is implemented to address the shortcomings de biotechnologie Strasbourg in 2008, a of each method, and to identify deletion- trinational biotech engineering school in the duplication events, which represent the heart of Europe. most common form of superposed SV. We have developed a streamlined visualization After that he did a PhD in Bioinformatics procedure, allowing inspection of SV in affiliation with the Max Planck Institute with their underlying evidence in genome for Molecular Genetics, Berlin, Germany browsers. We have performed extensive and the Centre for Genomic Regulation, benchmarking against the cell line NA12878 Barcelona, Spain. During his PhD, Andre and obtained sensitivity and precision up to assembled complex genomes de novo, 90% for deletions between 1-10kb. Compared annotated them and investigated the to microarrays currently used in clinical genome sequence variability and genome genetic testing to detect copy number events evolution. >50kb, WGS detects 25% more gold standard He completed his PhD with magna cum deletions. Additionally, we detect 4,000 CNVs laude in 2013, published his main research below the resolution of microarrays as well as work in Nature, and joined the Garvan 144 copy neutral deletion duplication events Institute in 2014. and 17 inversions. We present our findings from applying this workflow to ~100 whole Abstract genomes from clinical samples, including a Genomic structural variations (SVs), including cohort of 60 dilated cardiomyopathy patients. copy number variants (CNVs) affect 100 fold more nucleotides and arise de novo 1,000- We conclude that WGS is superior to cur- fold more frequently than single nucleotide rent best-practice use of microarrays in the variants (SNVs). SVs are recognised to clinic in terms of both analytic sensitivity and cause many complex developmental, neuro- precision. cognitive and neuro-behavioural phenotypes. Although microarrays have been the mainstay of diagnosing CNVs in patients, the broad

42 2015 AGTA Conference

1235-1250 has modified and enhanced the popular LOVD3 to create a tool MG-LOVD. MG- SHARED ANALYSIS FOR GENOME LOVD combines the functionality of locus TESTING IN THE HEALTH CARE SETTING specific and public variant databases with clinical diagnostic analysis workflows for DR NATALIE THORNE variant curation, interpretation and reporting Melbourne Genomics Health Alliance, VIC in a multi-institutional setting. Modifications BIOGRAPHY have also been made to allow variant data Natalie Thorne is the project manager for sharing among a local network of instances clinical bioinformatics and genomics in the within the Alliance. This is an initial step Melbourne Genomics Health Alliance. After towards sharing via a single data repository. a degree majoring in mathematics and An overview of how Cpipe and MG-LOVD statistics and minoring in genetics, Natalie are implemented and used in the Alliance completed her PhD at the Walter and Eliza will be presented. The key challenges and Hall Institute’s Bioinformatics division in benefits arising from sharing a common 2004. She worked for Cancer Research pipeline and database will also be dis- UK at the Cambridge Research Institute for cussed. nearly 5 years before returning to the Walter and Eliza Hall Institute. In her current role * http://www.genomemedicine.com/con- in the Melbourne Genomics Health Alliance, tent/7/1/68 she has the unique experience of managing clinical and diagnostic bioinformatics 1250-1305 aspects across 10 organisations who are collaboratively working to implement BRAIN-cX – AN INTERACTIVE WEB-TOOL genomic testing into the health care system FOR GENE PRIORITISATION AND MORE in Victoria. SASKIA FREYTAG Abstract The Walter + Eliza Hall Institute of Medical Enormous efforts have been put into Research, VIC translating genomic tests into the clinical BIOGRAPHY setting, particularly in major research, Saskia studied Statistical Science at UCL in medical and diagnostic centres across London. After finishing she moved to Ger- the world. To achieve this, change in the many to pursue a PhD in Biostatistics focus- health care system and collaboration is ing on GWAS analysis. In 2014 she moved needed. The ten members of the Melbourne to Melbourne where her research focus is Genomics Health Alliance are working mircoarray cleaning, co-expression analysis together to implement common systems and the development of interactive analysis and standards for genomics for Victoria. tools. These include a common bioinformatics pipeline, variant curation database and Abstract genome data repository. The Alliance Whole-exome sequencing is a valuable pipeline, Cpipe (Cpipeline.org *), now clinical tool for human disease. However, being utilised by member organisations, for many sporadic patients this approach is undergoing accreditation in two yields hundreds of credible mutations that laboratories. Development is progressing could be the cause of the patients’ disorder. in a collaborative manner and driving Sifting through these mutations individually standardisation. In addition, the Alliance is a time-consuming and taxing process and

43 2015 AGTA Conference

often fails to establish a feasible list of fol- low-up candidates. Gene prioritisation meth- ods promise to identify the most interesting mutations located in genes using computa- SESSION 6 tional methods, which are ideally applied to tissue-specific genomic data. Unfortunately, results from these methods are not usually readily available to clinicians, as the gener- GENOMICS OF PLANTS AND ation of such methods, and often their inter- FINE WINE pretation, requires trained bioinformaticians. We have developed an intuitive web-tool for gene prioritisation in neurological disorders: Chaired by Professor Justin Borevitz and brain-coX. Dr Rose Andrew brain-coX incorporates and combines six 1515-1555 large datasets on gene expression in the developing and ageing human brain. These ORIGIN AND CONSEQUENCES OF datasets are adaptively cleaned of unwant- GENETIC AND EPIGENETIC VARIATION ed variation, maximising information for the IN ARABIDOPSIS THALIANA AND ITS disease of interest. In addition to gene pri- RELATIVES oritisation brain-coX’s functionality includes PROFESSOR DETLEF WEIGEL extensive network visualization options as Max Planck Institute for Developmental well as interactive analysis tools to explore Biology, Germany changes in the gene-gene network along brain development. In the future, we hope BIOGRAPHY to extend brain-coX offering interactive tools Detlef Weigel is a German-American that visualize differences between brain scientist. He studied biology and chemistry regions. in Bielefeld and Cologne, and received a PhD from the University in Tübingen in Currently, our clinical collaborators and we 1988. For his postdoctoral work at Caltech, are extensively testing brain-coX to prioritize he switched from Drosophila to plants. He candidate genes for patients with childhood joined the faculty of the Salk Institute in La epilepsies. We find that the use of brain-coX Jolla in 1993, and has been a director at not only empowers clinicians and biologists the Max Planck Institute for Developmental with no programming or significant statistical Biology since 2002. The work of Detlef knowledge, but also leads to better commu- and his group has been recognized by nication between collaborators. several awards, including the Gottfried brain-coX is available via shiny.bioinf.wehi. Wilhelm Leibniz Award of the Deutsche edu.au/freytag.s Forschungsgemeinschaft. He is a member of the US National Academy of Sciences, the German National Academy of Sciences Leopoldina and the Royal Society.

The first major finding from the Weigel lab was that a gene from Arabidopsis thaliana could dramatically accelerate the flowering of trees; this established a proof of concept

44 2015 AGTA Conference

for Arabidopsis genetics as a platform for Abstract biotechnological discoveries. His group later My group is addressing fundamental discovered the first plant microRNA mutant, questions in evolutionary biology that inform and they identified the factor that is now also applications in breeding: (i) Where do known to be the long sought-after mobile new genetic variants come from? (ii) Why flower-inducing signal. In collaboration with do some variants increase in frequency? his Salk colleague Joanne Chory, Detlef was (iii) And why are some combinations of new one of the first to exploit Arabidopsis natural variants incompatible with each other? genetic variation for understanding how the environment affects plant development. In collaboration with other labs (Bergelson, In recent years, his work in the areas of Ecker, Mott, Nordborg & K. Schmid), and with evolutionary genetics and genomics has the help of Monsanto, we have been describing focused on plant immunity and epigenetics. whole-genome variation in natural accessions In addition to hypothesis-driven research, of A. thaliana (http://1001genomes.org). the Weigel group has a long history of In addition to resequencing efforts, this providing new technologies and resources to increasingly includes de novo assemblies. To the community. They have pioneered several better understand the patterns we observe in new forward genetic methods, and produced A. thaliana, we are comparing within-species the first haplotype map outside of mammals, variation with differences to the closest which put Arabidopsis at the forefront of relative, A. lyrata, and to variation in a related genome-wide association studies. This has genus, Capsella (with Neuffer, Nordborg & culminated in an effort to sequence the Wright labs). genomes of 1001 natural A. thaliana strains On the other end of the spectrum, we are (The 1001 Genomes Project). analyzing new DNA mutations and epigenetic Detlef has an extensive record of service variants that have arisen spontaneously under to the scientific community. He has been a laboratory conditions or in a natural mutation co-organizer of over a dozen international accumulation experiment. The latter studies conferences, and he has served on several take advantage of an A. thaliana lineage that editorial boards as well as advisory boards was apparently introduced to North America of biotech firms and research institutes. He in historic times and accounts for about half is also a forceful advocate of open access the population there (with Bergelson and publishing. He was among the initial editorial Burbano labs). We have been able to support board of PLoS Biology, and is now Deputy what we see in the extant North American Editor of eLife, the new high-level journal population by whole-genome sequencing of th supported by HHMI, Wellcome Trust and herbarium samples from the 19 century. Max Planck. He is a co-founder of three The ultimate goal of our top-down studies is to startups in the biotechnology arena, most understand how new genetic and epigenetic recently Computomics, which provides variation interact with reassortment of variants bioinformatics services, and CeMeT, which after crosses and natural selection to shape provides human metagenome analyses. geographic patterns of diversity. To this end, Detlef would be particularly keen to discuss we are following natural populations during publication strategies, organization of the season and over consecutive years. This scientific meetings, social media outreach work in turn is complemented by forward and new technologies in genomics. genetic analyses, especially of detrimental combinations of sequence variants found in separate lineages (with Dangl lab).

45 2015 AGTA Conference

Additional information about our work can be Rob is keen to discuss applications of marker found on our website, http://weigelworld.org. technologies, managing and analyzing large data sets and potential collaboration 1555-1630 opportunities with like-minded scientists.

DEMOCRATISING GENETIC ANALYSIS Abstract AND BREEDING WITH GENOTYPING BY Genotyping-by-Sequencing (GBS) is a high SEQUENCING throughput method for generating many thousands of genetic markers using next

MR ROB ELSHIRE generation sequencing. The method was The Elshire Group Limited, New Zealand developed with simplicity of deployment BIOGRAPHY and an open source philosophy in mind. Rob is a Principal Scientist with The Elshire We published the method in 2011 in an Group Limited with a long history working open access journal so that as many with molecular markers in agriculture. Rob researchers as possible could benefit from was the inaugural director of the Illinois its use and potentially modify or extend it. Genetic Marker Center at the University of The method has been adopted by many Illinois. This facility provided the agricultural researchers. In addition to plant breeders research community there access to in commodity species, GBS is being used molecular marker technology in a cost by breeders of orphan crops as well as in effective manner. At Cornell University, the ecological sciences. Simplicity, low cost he developed a low cost, high-throughput and openness of the method have allowed genotyping method commonly known as many researchers to jump into the genomics genotyping-by-sequencing (GBS). While the era and leveled the research playing field. impetus for developing GBS was to serve However, more than a good molecular the needs of the plant breeding community, method is needed to democratise genetic the technology has enjoyed wide adoption in analysis. many areas of biological research. The Biospectra by Sequencing project aims In addition to developing the molecular to develop a set of community resources. method, he has collaborated in transnational There are two core components. One is research projects, organized numerous an interactive information repository to workshops and contributed code to analyze contain useful information that is not often GBS data in a reproducible way. He published in the literature. The other is a continues these efforts with his current work Free / Libre Open Source Software (FLOSS) which focuses on collaboratively building project developing the software tools for New Zealand capability in GBS and related best practice in data quality, automation technologies to enable genetic analysis and reproducibility. Working together, we in a wide range of NZ specific biological will reduce duplication of efforts, provide questions. a high quality set of information and tools and further reduce the barriers to entry Rob is an enthusiastic supporter of Free / in genomics research. We aim to enable Libre and Open Source Software and was researchers to do better research at a lower presented the New Zealand Open Source price than previously possible. People’s Choice award in 2014 for his genotyping-by-sequencing contributions to the community.

46 2015 AGTA Conference

and assembled using ALLPATHS-lg, with further scaffolding carried out by SSPACE2. 1630-1645 As expected, assembled scaffolds represented the metagenomic content of the UNRAVELLING THE MāNUKA GENOME: A field-grown mānuka. Several approaches METAGENOMIC APPROACH were taken to differentiate the various MS AMALI THRIMAWITHANA fractions of the metagenome. The final The New Zealand Institute for Plant & Food assembly of the plant “partition” resulted Research Limited, New Zealand in an assembly containing 2505 scaffolds representing 298Mb, with an N50 of 262Kb BIOGRAPHY and an overall CEGMA score of 94%. Amali is a Bioinformatician at the New Gene prediction (aided by RNAseq data Zealand Institute for Plant and Food and expressed sequence tags (ESTs) from Research Limited since 2011, after Eucalyptus grandis) on the plant partition graduating with a Masters of Bioinformatics scaffolds predicted 43,716 genes. The from University of Auckland. She has been resulting gene predictions were then used to involved in research projects focusing on carry-out comparative genomics to explore fungal pathogens, insect pests and plants the gene space of mānuka further. where she has undertaken bioinformatics analysis including de novo assembly and 1645-1700 annotation of genomes and transcriptomes, THE SMRT WAY TO SEQUENCE A YEAST differential expression as well as GENOME comparative genomics. Her current major areas of interest focus on understanding DR RICHARD EDWARDS insect pests of important horticultural crops University of New South Wales, NSW as well as genome exploration of native New Zealand plant species. BIOGRAPHY Rich Edwards received his undergraduate Abstract and postgraduate training in genetics at Mānuka (Leptospermum scoparium), is a the University of Nottingham in the UK. He shrub indigenous to eastern Australia and moved to Dublin in 2001 for a postdoctoral New Zealand. The species is well known for position in bioinformatics before returning to its phenotypic plasticity, resulting in a range the UK in 2007 to establish his own research of chemotypes. This plasticity has resulted group at the University of Southampton. Rich in a wide range of products fromthis shrub, moved to UNSW in 2013 as a senior lecturer including mānuka honey, essential oils in Bioinformatics. and pharmaceutical products with unique antimicrobial activity. Currently, fundamental Research interests in the Edwards lab stem knowledge on the molecular biology of from a fascination with molecular basis of mānuka is minimal and hampered by a lack evolutionary change and how to harness of availability of a whole genome sequence the genetic sequence patterns left behind to for the species. Acquisition of a whole make useful predictions about contemporary genome sequences is complicated by the biological systems. Core research in the lab presence of endophytic fungi, with mānuka is the study of Short Linear Motifs (SLiMs), commonly associated with one or more which are short regions of proteins that endophytic fungal species in the wild. DNA mediate interactions with other proteins. from field-grown mānuka was sequenced Rich is also excited about the new

47 2015 AGTA Conference

opportunities for functional genomics that the assembly process and derive assembly the new generation of long read sequencing settings for two novel strains. To this end, technologies are making possible. we have developed a new pipeline for the comparative assessment of high quality Abstract whole genomes against a reference. PacBio Single Molecule Real Time (SMRT) Genome assembly using the PacBio SMRT sequencing is rapidly becoming the Portal is a two-step process, with HGAP technology of choice for de novo whole generating a “pre-assembly” of error- genome sequencing. The long read lengths corrected “seed” reads that are subsequently and random error of PacBio data should assembled using the Celera assembler. make genome assembly considerably We explore the trade-off between accuracy easier and more accurate than Illumina and sequencing depth of this pre-assembly short read data. Indeed, the hope is that for different seed read length cutoffs and adoption of this technology will enable how this affects the final assembly. The highly accurate and complete genome effect of DNA loading concentration will sequences to be generated in the absence also be discussed. Optimised assemblies of either a reference genome or extensive will also be compared to results from the bioinformatics support. developmental FALCON diploid assembler. Here, we report on the de novo genome 1700-1730 assembly of three yeast genomes from SMRT data generated using the new WINE ‘OMICS: AT THE CUTTING-EDGE PacBio RSII at the UNSW Ramaciotti OF THE OLDEST BIOTECHNOLOGY Centre for Genomics. A haploid reference yeast genome strain, S288C, and two novel DR ANTHONY BORNEMAN diploid strains were sequenced as part of Australian Wine Research Institute, SA a larger functional genomics project. For each strain, 20kb SMRT Bell library preps BIOGRAPHY Anthony obtained his PhD from the were performed and sequenced on two University of Melbourne where he studied SMRT Cells using the P6-C4 chemistry. the regulation of morphology in the fungal Between 1.74 Gb and 2.55 Gb of usable pathogen Penicillium marneffei and spent sequence data was generated for each four years as a postdoctoral associate with strain, with read lengths of up to 53.3 kb. Prof. Michael Snyder at Yale University, Whole genome de novo assemblies are then applying whole-genome techniques to generated through the PacBio SMRT Portal. compare transcriptional networks across We are using the S288C data to explore yeast species. Anthony is currently a performance in comparison to the Principle Research Scientist at the Australian published genome as a reference. An Wine Research Institute. His research is initial assembly of S288C yielded over focused on applying genomics, systems- 99.9% genome coverage at 99.997% and synthetic-biology to understand the accuracy on only 29 unitigs, versus 17 genetic basis of phenotypic diversity in reference chromosomes (16 nuclear industrial microorganisms, with particular chromosomes plus mitochondrion). Of focus on the wine yeasts Saccharomyces these, 15 chromosomes were essentially cerevisiae and Brettanomyces bruxellensis. returned as a single, complete unitig. We are now using the S288C data to optimise

48 2015 AGTA Conference

Abstract Wine is arguably the oldest biotechnological WEDNESDAY endeavour, with evidence of winemaking 14 October 2015 dating back at least 7,000 years. Despite its artisan nature, pioneering scientists such as Lavoiser and Pasteur positioned wine research at the cutting-edge of the biological and chemical sciences, a position it still SESSION 7 holds to this day. In a modern day context, the application of technologies such as genomics, metagenomics, systems biology and metabolomics are revolutionizing wine QUANTITATIVE GENETICS AND research by yielding insights into the true DECODING THE GENOME breadth of genetic diversity that exists across the varied biological inputs that are central to this age-old industry and how this Chaired by Dr Helen Speirs and diversity translates into metabolomic, and Dr Alexandra Livernois therefore taste and aroma, impacts on the finished wine. 0900-0930 CAN WE EXPLORE THE GENOMICS OF CANINE WORKING BEHAVIOUR WITH SELECTIVE SWEEP ANALYSIS?

PROFESSOR CLAIRE WADE Faculty of Veterinary Science, The University of Sydney, NSW

BIOGRAPHY Claire Wade began her career in quantitative genetics before making the leap to genomics in 2002 when she began a position with the Whitehead Institute for Biomedical Research at Massachusetts Institute of Technology. The genomics group at the Whitehead later became one of the founding groups of what is now the Broad Institute. While in the USA, Claire worked on several mammalian genome projects including the mouse, dog and horse (for which she was the lead researcher).

Claire’s research interests include unravelling the secrets of genome biology using next generation sequencing. In particular, she studies the application of new genomic technologies to improve our understanding of diseases and behavioural

49 2015 AGTA Conference

traits in domestic animals and wildlife and selective sweep analysis to determine regions our understanding of the links between DNA that might be of importance in Guide Dog and phenotype in general. success in the NSW Guide Dog Association. Our analyses reveal strong divergent sweep Projects currently underway are as diverse signals in both of these breed groups. as studying the genetics of durability in Thoroughbred race horses, finding genes 0930-0945 underlying canine separation anxiety and working dog performance, improving captive PROGRAMMABLE RNA TARGETING AND animal management using new genetic CLEAVAGE BY CRISPR/CAS9 resources, and developing new methods for the computational analysis of high throughput DR MITCHELL O’CONNELL University of California, Berkeley, USA sequencing data. Other projects involve mapping genes causing congenital disorders BIOGRAPHY in dogs including cleft palate and deafness Mitchell received his PhD in Biochemistry using whole genome association analysis and from the University of Sydney in 2013, genotyping by sequencing. where he worked in Prof. Joel Mackay’s group to develop artificial zinc-finger ABSTRACT proteins to specifically recognize RNA Selective sweeps are genomic signatures for sequence-specific RNA-targeting of human or natural intervention in animal applications. In 2013, Mitchell moved to fitness. The predominant feature of a selective the University of California, Berkeley to sweep is that there is a long region in the take up a postdoctoral position in the lab DNA that has little remaining variation in the of Prof. Jennifer Doudna. In the Doudna cohort of animals under selection. While most lab, Mitchell has focused on exploring the studies have in the past compared widely molecular mechanisms of RNA-mediated different dog breeds, that are expected to gene regulation and the development of exhibit divergence over a large portion of tools to interrogate these processes. In early the genome, we have been concentrating 2012, the Doudna laboratory was the first to our analyses within single dog breeds that show the programmable, sequence-specific have been subjected either to a formalised DNA recognition and cleavage ability of a breed split, or to very different selective CRISPR-associated protein known as Cas9. pressures. Here, we focus on selection for This research led to an explosion in the working behaviour in two breed groups. use of CRISPR/Cas9 as a genome-editing The Kelpie is a breed that was developed in tool in multiple cell types and organisms. Australia for working stock (predominantly However, this protein had been thought sheep and cattle). Pedigrees for Kelpies are to be incapable of targeting RNA. In a maintained by two registration organisations, recent publication in Nature, Mitchell and one that focuses on working ability, and the colleagues were able to show that this is other that focuses on companionship and not the case and Cas9 can be used as a conformation. Dogs are rarely if ever mixed readily programmable sequence-specific between these registries in Australia. New RNA-binding and cleavage enzyme. This South Wales Guide dogs are derived from discovery paves the way for sequence- a mixture of Golden Retriever and Labrador specific tools to explore the myriad of RNA breeds, but by genomic clustering the dogs species implicated in gene regulation and are predominantly Labrador Retriever at disease. Mitchell is now building on this work the genomic level. We have employed a by exploring a number of applications of

50 2015 AGTA Conference

CRISPR/Cas9 for untagged RNA transcript 0945-1000 analysis, detection and manipulation. EXPRESSION QUANTITATIVE TRAIT ABSTRACT LOCI (eQTLs) OF ENDOMETRIOSIS RISK The CRISPR-associated protein Cas9 is an LOCUS AT 1p36.12 RNA-guided DNA endonuclease that uses RNA–DNA complementarity to identify target DR JENNY FUNG sites for sequence-specific double-stranded QIMR Berghofer Medical Research Institute, DNA (dsDNA) cleavage. CRISPR/Cas9 has QLD proven to be a versatile tool for genome BIOGRAPHY engineering and gene regulation in a large Dr Jenny Fung obtained her PhD from the range of prokaryotic and eukaryotic cell types, University of Queensland in 2013. Under and in whole organisms, but it has been the supervision of Prof. Chen Chen and A/ thought to be incapable of targeting RNA. I Prof. Lisa Chopin with a project aimed at will present our research showing that Cas9 investigating the underlying mechanisms for is able to bind with high affinity to single- endocrine hormones on endometrial cancer stranded RNA (ssRNA) targets in a guide-RNA cell progression by mRNA and protein programmable manner when an additional expression analyses on cancer cell lines short DNA oligonucleotide containing a and clinical samples, functional assays and required motif known as the protospacer lentiviral gene silencing in cancer cells and adjacent motif (PAM) is provided in trans. mouse xenograft models. In June 2013, Furthermore, we have shown that these Jenny took up a post-doctoral appointment PAM-presenting oligonucleotides (PAMmers) in Prof. Grant Montgomery’s Molecular stimulate site-specific endonucleolytic Epidemiology Laboratory to conduct large cleavage of ssRNA targets, similar to PAM- genetics and gene expression studies. mediated stimulation of Cas9-catalysed DNA Her research is focused on understanding cleavage. Using specially designed PAMmers, the complex mechanisms underlying the we have shown that Cas9 can be specifically associated genomic regions and identifying directed to bind or cut RNA targets while the important biological pathways increasing avoiding corresponding DNA sequences, and endometriosis risk by applying extensive we demonstrate that this strategy enables the functional studies and genetics and isolation of a specific endogenous messenger genomics statistical methods. RNA from cells. We envisage that Cas9 will enable new ways of investigating and Abstract manipulating many aspects of RNA function, Endometriosis is a common gynaecological including targeted degradation of RNAs, disease associated with chronic pelvic isolation and identification of RNA-associated pain, reduced fertility and dysmenorrhea. proteins, and in vivo imaging of RNAs. It affects 6-10% of women of reproductive age in Australia and is a complex disease influenced by multiple genetic and environmental factors. The causes and molecular mechanisms underlying the condition are largely unknown. We have used genome-wide association studies (GWAS) to understand the genetic architecture and discover genomic regions associated with risk for endometriosis.

51 2015 AGTA Conference

Although GWAS has been successful at finding common single nucleotide polymorphisms (SNPs) at risk loci, to enable translation of these results, a major SESSION 8 challenge remains to pinpoint the causal SNPs, identify their target genes and pathways contributing to endometriosis risk and to characterize their functional effects. MICROBIAL AND SINGLE CELL GENOMICS GWAS and replication studies have identified seven genomic regions with strong evidence for association with endometriosis risk. To Chaired by Dr Kirby Siemering and examine how genetic variants influence gene Dr Clare Stirzaker expression, expression Quantitative Trait Loci (eQTL) approach has been used, where 1045-1130 gene expression data and genetic variation data from the same individuals are integrated ILLUMINATING MICROBIAL DARK for statistical genetic analyses. Results from MATTER VIA SINGLE-CELL GENOMICS eQTL studies in peripheral blood samples DR CHRISTIAN RINKE from 962 individuals in the Brisbane System University of Queensland, QLD, Australia Genetics Study (BSGS) and a meta-analysis in world largest cis-eQTL study on peripheral BIOGRAPHY blood samples from ~5000 individuals Christian Rinke is a Research Officer at the identified two strong cis-associations Australian Centre for Ecogenomics (ACE), mapped to endometriosis risk loci at 1p36.12 University of Queensland, Australia. He with target genes, CDC42 (p= 9.8x10-198) received his PhD in Zoology from the Marine and LINC00339 (p = 5.01x10-53). The Biology Department at the University of target tissue for functional effects is not Vienna, Austria and has since shifted his known, but current theories suggest changes focus to the microbial world. in the endometrium. We are conducting studies of gene expression in samples of His research interests include genomics and endometrium in carriers of the risk alleles. the phylogeny and ecology of symbiotic and Integrating the GWAS findings and functional free living microbes. He focuses in particular characterization will provide a more on the uncultured majority of microbes (99%) comprehensive picture of the molecular links which elude current culturing efforts. This so between disease-associated variants and called “Microbial Dark Matter” can only be target genes and will greatly expand our explored with culture-independent methods. understanding of the mechanisms underlying Chris pioneered methods in high throughput the endometriosis susceptibility loci. single-cell genomics, the separation and sequencing of single bacterial and archaeal cells, and also employs metagenomics (the direct sequencing of environmental samples) to illuminate microbial dark matter.

ABSTRACT Our view of microbial genomic diversity is severely skewed with the majority of all sequenced bacterial and archaeal genomes

52 2015 AGTA Conference belonging to only four bacterial phyla. This holds a B.Sc. degree & Ph.D. He was bias results in part from our inability to awarded postdoctoral clinical and research cultivate most microbes, a necessary step fellowships in the UK (Queen’s University for traditional whole genome sequencing. Belfast) and USA (University of Chicago) Through cultivation-independent approaches and has held travel fellowships in South such as single-cell genomics, one can now America, and Switzerland, and an Honorary explore the genetic diversity and metabolic University Fellowship in the UK (University potential of uncultivated environmental of Manchester). In addition, he won a microorganisms. We successfully amplified Silicon Valley award for innovative web several hundred single cells from free-living design, published numerous, per reviewed, and symbiotic populations without cultured original research papers and has 4 assigned representatives, known as microbial dark patents. matter. The single-cell genomes allowed us to explore their intra- and inter-phylum-level ABSTRACT relationships, to decipher encoded pathways, Fluorescent activated cell sorting (FACS) and to discover novel metabolic features. The has the ability to select, characterize and single- cell reference genomes also facilitate quickly deposit single cell(s). It does this with the interpretation of metagenomic data sets high precision into PCR plates that can be and substantially improve phylogenetic preloaded with reagents suitable for further anchoring of up to 20% of metagenomic reads downstream genomic analysis. It also has in some habitats. While there is still much the ability to connect cell phenotype with ground to cover, single-cell-genomics has genomic data. Here we describe two novel proven to be a valuable tool to improve our FACS innovations – detection through total understanding of microbial evolution on earth. internal reflection, and automated droplet monitoring, that brings the power of flow 1130-1150 cytometry within the reach of individual genomic labs. HIGH THROUGHPUT CELL SORTING FOR DOWNSTREAM GENOMIC ANALYSIS 1150-1205 DR J CLARK MASON SINGLE-CELL ANALYSIS OF BREAST BD Biosciences, USA CANCER MOLECULAR SUBTYPE REVEALS CLINICALLY RELEVANT HETEROGENEITY

MS LAURA BAKER Garvan Institute of Medical Research, NSW

BIOGRAPHY BIOGRAPHY Dr J. Clark Mason is Senior Director, I’m a PhD student at the Kinghorn Cancer Genomics Innovation with BD Biosciences, Centre (Garvan Institute of Medical where his focus is strategic platform Research). My research focuses on the development for the cell analysis market. intra-tumoural heterogeneity observed He came to BD with more than 15 years’ clinically in the diagnosis and treatment of experience as a marketing executive, breast cancer. We are further investigating bringing many key instrument and reagent what drives these fundamental cell platforms to the marketplace, for leading differences by using single-cell gene global life science companies. Clark expression techniques combined with breast

53 2015 AGTA Conference cancer subtyping analysis. This will enable predictor to examine single-cell intra-tumoral understanding of the clinically relevant heterogeneity of breast cancer subtypes single-cell heterogeneity that is masked across a range of models including breast when samples are analysed on a bulk cancer cell lines, patient derived xenografts tumour level. (PDX) and fresh patient tumours. Analysis of five breast cancer cell lines shows clear Another focus of my research is cellular gene expression diversity and single investigating a transcriptional regulator cells within a cell line possess different involved in mammary stem cell hierarchy PAM50 subtypes. Together these results and cellular differentiation as well as the suggest a high degree of heterogeneity regulation of poor prognosis basal-like within commonly studied breast cancer cell breast cancer. Unbiased proteomic (Rapid lines. This heterogeneity is an important Immunoprecipitation and Mass Spectrometry consideration when using models of breast of Endogenous proteins) and genomic cancer, as selected cell lines more accurately techniques (ChIP-exonuclease sequencing, recapitulate the complex heterogeneity ChIP-sequencing, RNA-sequencing) are observed using histological methods in yielding incredible insight into the molecular clinical samples. function of this protein. 1205-1220 ABSTRACT Breast cancers display significant clinical THE EXTREME MICROBIOME PROJECT and cellular intra-tumoral heterogeneity that (XMP): THE METAGENOMICS OF LAKE accounts for differential patient therapeutic HILLIER, A PINK HYPERSALINE LAKE responsiveness, local recurrence and metastasis. Molecular studies permit DR KEN MCGRATH detection of the five main molecular subtypes The Australian Genome Research Facility, of breast cancer through the introduction QLD of microarray technologies. The PAM50 subtype predictor uses a panel of 50 genes BIOGRAPHY Ken McGrath is the National Sequencing to predict molecular subtype using bulk Manager and Node Manager at AGRF patient RNA samples, however, this prevents Brisbane. He obtained his PhD in 2005 analysis of intra-tumoral heterogeneity and at the Uni. of Qld, and has a research simply measures average gene expression background in microbial community genetics, levels. In contrast, single-cell technologies including human and environmental enable analysis of gene expression variation microbiomes and metagenomics analysis. in individual cells and can molecularly inform Ken is currently involved with several on the wide variation in gene expression sequencing projects, including the US-based observed through standard histological extreme microbiome project (XMP), as well methods. Similarly, single-cell analysis of as evaluating emerging technologies such as molecular subtype may provide invaluable the Oxford Nanopore MinION information on clinically relevant tumour heterogeneity. Through this research, we ABSTRACT investigate whether the PAM50 signature of The Xtreme Microbiome Project (XMP) is a bulk tumour samples is simply an average of global scientific collaboration to characterize, diverse cellular states or if single cells can discover, and develop new pipelines and possess an independent PAM50 signature. protocols for extremophiles and novel In this study, we apply the PAM50 subtype organisms. The XMP has chosen a range

54 2015 AGTA Conference of extreme environments, including the oil-spilled deep sea brine lakes of the Gulf of Mexico, “The Door to Hell” gas crater in Turkmenistan, the public surfaces of the New York Subway System, and even the walls of the International Space Station.

As part of this project, the AGRF collected samples from Lake Hillier. This lake is a bright pink hypersaline lake located on Middle Island, part of the Recherche Archipelago in Western Australia. The constant colour of the lake is different to other pink lakes in the area, and is thought to be a result of unique microbial composition, including halophilic algae, archaea, and bacteria.

We analysed several sample types (lakeside bank sediment, surface water, submerged water, and lake sediment) using a wide variety of metagenomics techniques, including amplicon-based diversity profiling (16S V1- V3, 16S V3-V4, Archaeal 16S, Algal 18S, Fungal ITS; sequenced on MiSeq), shotgun metagenomics (HiSeq), metatranscriptomics (HiSeq), and long-read metagenomics (MinION). This data was compared to organisms confirmed to be present in the lake using culture-based techniques.

The resulting microbial profiles revealed distinct populations that inhabit the lakes various strata, and even hint at the historical misuse of the area. The analysis highlights the strengths and weaknesses of each method for microbiome analysis, as well as provides a detailed insight into the composition of an iconic feature of the Australian landscape.

55 ROCH1509C Roche Sequencing Banner A4_113354 HR.pdf 1 17/09/2015 10:03:13 AM

Roche Sequencing Changing Science. Changing Lives. Pioneers Legacy of fi rsts with a focus on innovative solutions Value Value through complete sequencing solutions that change the way people use science and medicine Certainty A trusted partner that provides quality products and support— reliable, sustainable, fl exible Breadth Scientifi c discovery to clinical validation with global reach

454, 454 LIFE SCIENCES, NIMBLEGEN, and SEQCAP are trademarks of Roche. For life science research only. Not for use in diagnostics. © 2015 Roche Diagnostics. 2015 AGTA Conference

POSTERS 2015 AGTA Conference

POSTER PRESENTATIONS Monday 12 October 2015

PRESENTING # POSTER TITLE AUTHOR A Bioinformatics Analysis of the Transcription Factors 1 Associated with the Interferon Response in Human and Dr Ross Chapman Mouse Models RNA-Seq: What difference does a strand-specific protocol 2 Dr Susan Corley make?

Host microRNAs are promising biomarkers of infectious Dr Christopher 3 diseases Cowled Whole Genome Sequencing Identifies Structural Variation 4 and Non-Coding Variants in the Region Overlapping the Dr Alexander Drew CMTX3 and CMTX4 Loci De novo trancriptome assembly and identification of 5 Dr Ali Livernois cytokine genes in Pogona vitticeps The Novel Long Non-Coding RNA RP1X Promotes Mr Bernard 6 Neuroblastoma Cell Proliferation by Stabilizing N-Myc Atmadibrata Protein DNA Methylation in Patients with a Strong Family History Professor Vicky 7 of Early Onset Coronary Heart Disease Cameron Using population sequencing data to investigate the 8 evolutionary role and functional impact of inversion Mr Haojing Shao polymorphisms Whole Genome vs Targeted Sequencing: Which Provides 9 Mr Mark Pinese Better Diagnostic Yield for Disease-Causing Variants? Bulked Segregant - genotyping-by-sequencing: Cost- Mr Kokulapalan 10 effective and background independent genetic mapping of Wimalanathan mutants and QTL Maize - GO Annotation Methods Evaluation and Review Mr Kokulapalan 11 (Maize-GAMER) Wimalanathan

Genetic Characterisation of the Evolution of a Novel Dr Åsa Pérez- 12 Metabolic Function in Yeast Bercoff

Early Career Student Researcher

58 2015 AGTA Conference

PRESENTING # POSTER TITLE AUTHOR An immune transcriptional regulatory network approach to Dr Margaret 13 Multiple Sclerosis Jordan “Profiling the profilers”: Evaluating current bioinformatics Mr Naga 14 methods for characterising the diversity of microbial Kasinadhuni communities Human Whole Genome Sequencing with Low and 15 Dr Kirby Siemering Ultralow DNA Inputs

The Expression of Small Nucleolar RNAs in Human 16 Mr Hyun Jae Lee Tissues

Evaluating the efficiency and reproducibility of genotyping 17 Dr Rust Turakulov by sequencing.

Quantification of cellular features using novel image 18 Mr Matloob Khushi analysis algorithms Non-coding mutations in cancer alter G-quadruplex 19 stability and affect its regulatory role in the 5’ untranslated Mr Mahdi Zeraati region of mRNA Small ribosomal subunit profiling quantifies conformational 20 Dr Stuart Archer changes in initiating and terminating ribosomes.

Cancer gene-interaction analysis using a data-driven 21 Dr Denis Bauer inverse covariance matrix approach

Functionally characterizing every single mutation in Your 22 Dr Alan Rubin Favorite Gene Promiscuous DNA-binding of a mutant zinc finger Professor Andrew 23 protein corrupts the transcriptome and derails erythroid Perkins differentiation Characterisation of novel hypomorphic and null mutations Professor Andrew 24 in Klf1 derived from a genetic screen for modifiers of Perkins alpha-globin transgene variegation

59 2015 AGTA Conference

POSTER PRESENTATIONS Tuesday 13 October 2015

PRESENTING # POSTER TITLE AUTHOR Elucidating the function of the Evx1 protein coding/ Professor Andrew 25 lncRNA sense-antisense locus during gastrulation Perkins Two Orthogonal Processes Underlie Survival of Patients 26 with Resected Pancreatic Cancer: A Successful Mr Mark Pinese Application of Expression Deconvolution Discovering genomic hotspots for columnar growth In 27 Ms Cecilia Deng Malus X domestica with GBS

Gene Ontology enrichment analysis for gene subsets of 28 Dr Denis Bauer distinctive function

The effect of ADAR3 deficiency on RNA editing in Dr Dessislava 29 mouse hippocampus Mladenova

Harnessing the power of serum microRNAs to predict 30 Mr Jaynish Shah surgical outcome for women with ovarian cancer

Application of genome wide analysis in developmental 31 Dr Ivan Prokudin eye disease

Disentangling Genetics and Methylation in the Major Dr Jovana 32 Histocompatibility Complex Maksimovic A Novel Variable Short Tandem Repeat in the Upstream 33 Regulatory Region of the Estrogen-Induced Gene Mrs Katherine Bolton EIG121 is Potentially Involved in Cancer Risk Deleterious Passenger Mutations as a Marker for Ms Magdalena 34 Progression towards Liver Cancer Budzinska

Loss-of-function germline FGFR1 mutation identified in a 35 Dr Mark McCabe patient with prolactinoma Simultaneous “multi-omic” measurement of gene 36 fusions, mRNA and proteins at 800-plex using single- Dr Michael Rhodes molecule optical barcodes. Wild Wine: metagenomic analysis of microbial 37 Dr Peter Sternes communities during wine fermentation.

60 2015 AGTA Conference

PRESENTING # POSTER TITLE AUTHOR High quality RNA isolation from rat spinal cord motor 38 Dr Prachi Mehta neurons using Laser Capture Microdissection

Impact of Gadolinium Oxide Nanoparticles on Viability 39 Dr Saad Alkahtani of Human Liver-Derived Cell Line

Genomic insights into the nitrate assimilation potential 40 Mr Ryan Zeppel of Brettanomyces bruxellensis isolates. Epithelial, metabolic and innate immunity 41 transcriptomic signatures differentiating the rumen from Dr Ruidong Xiang other sheep gastrointestinal tract tissues ROS Mediated Apoptosis and DNA Damage Induced Professor Saud 42 by Barium Nanoparticles In Mouse Embryonic Alarifi Fibroblasts Seave: a Comprehensive Variant Filtration Platform for Dr Velimir 43 Clinical Genomics Gayevskiy

DEAR-O: Differential Expression Analysis based on Ms Zong-Hong 44 RNA-seq data -Online Zhang Analysis of Metastasis Progression in High Grade Dr Matthew 45 Mucinous Ovarian Cancer by Whole Genome Wakefield Sequencing of Multiple Autopsy Sites Functional Characterisation of Human Pseudogene 46 Dr Daniel Thomson Transcription using Targeted RNA Sequencing The presence of the rs17878362 polymorphism in Dr Kelly Avery- 47 breast cancer is associated with a low Delta-40p53:p53 Kiejda ratio and outcome. A High Density SNP-based Genetic Linkage Map of 48 Sweet Potato Established by Scaffolding an Ipomoea Mr Chenxi Zhou Trifida (H. B. K.) G. Don. de novo Assembly A novel technology for whole genome amplification 49 Ms Martha Hinton from single cells and limited material A single-tube NGS library prep workflow integrating 50 enzymatic fragmentation results in high yields and low Ms Martha Hinton sequence bias

51 Cooking up a pathway analysis with roast and fry Dr Goknur Giner

61 2015 AGTA Conference

among genes that were regulated by MONDAY interferon in both human and mouse. Many transcription factors were common to both 12 October 2015 species, indicating that the regulation of the genetic response is highly conserved between the two species. However, notable POSTER 1 differences between the species were detected for some transcription factors, A BIOINFORMATICS ANALYSIS OF THE indicating that some divergent evolution TRANSCRIPTION FACTORS ASSOCIATED in the interferon response might have WITH THE INTERFERON RESPONSE IN occurred between the two species. Potential HUMAN AND MOUSE MODELS inter-species diversity in the regulation Ross Chapman, Helen Cumming, Linden of the interferon response is worthy of Gearing, Paul Hertzog experimental validation.

Abstract POSTER 2 Interferons play a vital role in mammalian genetic responses to bacterial and viral RNA-SEQ: WHAT DIFFERENCE DOES A infections and to tumour development. STRAND-SPECIFIC PROTOCOL MAKE? The interferon response following disease Susan Corley, Zara Ali, Karen MacKenzie, challenge is highly modulated; the host Marc Wilkins response must be sufficient to protect the host, while excessive responses can cause Abstract toxicity and even death. Furthermore, RNA-Seq technology is capable of the interferon response follows a highly detecting all forms of RNA transcribed from structured temporal response, with the genome (including mRNA coding for stimulation precipitating an organised proteins, microRNA, snoRNA, lincRNA, sequence of transcriptomic events. and mtRNA). Until recently RNA-Seq Transcription factors play an important role experiments were generally conducted in the regulation of genetic transcription, using a single stranded protocol. Using influencing the timing and intensity of this protocol it was not possible to detect expression of individual genes. Here we whether transcription had occurred from the present a study of the relationship between forward or reverse strand of DNA. However, transcription factors and the genetic recent developments in sequencing responses to interferon stimulation in human chemistry have made it feasible to routinely and mouse model systems, conducted using identify the DNA strand from which any RNA two bioinformatics tools: 1) the Interferome transcript has originated (strand-specific on-line database of interferon-regulated protocol). Here we compare the outcomes genes (http://interferome.its.monash.edu. of read mapping and feature counting using au/ interferome/home.jspx), and 2) the both a non-stranded and a strand-specific Enrich custom software, which is powerful protocol. Our data is derived from 2 studies and intuitive tool for revealing transcription using human primary hematopoietic cells. factors involved in the regulation of co- Sequencing was performed using the expressed genes. Illumina NextSeq platform and the RNA libraries were prepared using Illumina’s The analysis revealed populations of TruSeq Stranded Total RNA kit with Ribozero transcriptions factors that were enriched pulldown. Over 40 million 75 bp paired-end

62 2015 AGTA Conference reads were produced per sample. Reads (no splicing and moderate diversity), and were mapped to the human reference each sequencing read corresponds to (Ensembl GRCh37) using Tophat2 and exactly one count (no need to adjust for reads mapping to features were counted transcript length). Despite these advantages, using HTSeq. Here we present a breakdown difficulties still remain. In particular, of the types of RNA, coding and non-coding microRNA quantification is somewhat RNAs detected using the strand-specific and platform-dependent. Reasons for this non-strand specific protocols and compare include difficulty with normalization due to the outcomes of differential expression skewed sample composition (the single most analysis using these protocols. abundant miRNA can make up > 95% of some samples), and the natural occurrence POSTER 3 of isomiRs can influence abundance estimates. Nevertheless, host microRNAs HOST MICRORNAS ARE PROMISING are a promising class of infectious disease BIOMARKERS OF INFECTIOUS biomarkers. DISEASES

Christopher Cowled, Cameron Stewart, POSTER 4 Chwan-Hong Foo, Andrew Bean WHOLE GENOME SEQUENCING ABSTRACT IDENTIFIES STRUCTURAL VARIATION Molecular diagnostic tests for infectious AND NON-CODING VARIANTS IN THE diseases may be either direct or indirect. REGION OVERLAPPING THE CMTX3 AND Direct tests target pathogen-derived CMTX4 LOCI molecules (such as protein antigens or Alexander P. Drew, Megan H. Brewer, nucleic acids), while indirect tests target Andrzej Kochański, Aditi Kidambi, Gonzalo host-derived molecules, such as antibodies. Pérez-Siles, Adrienne Grant, Joanna Direct tests can face problems with fast- Kosińska, Dagmara Kabzińska, Rafal Płoski, evolving pathogens such as viruses, or Irena Hausmanowa-Petrusewicz, and Garth highly localized infections, or agents that A. Nicholson and Marina L. Kennerson replicate at very low levels. In contrast, antibodies are generally undetectable for the ABSTRACT first week of infection, frequently cross react Previously we reported a family with a against related pathogens, and cannot easily Mendelian inherited, axonal Charcot Marie distinguish current from past infections. Tooth disease, which is genetically unsolved There is therefore a constant need for new (Kochanski, Kennerson et al. Neurology and better diagnostic tests. Quantitative 2005; 64:533-535). We performed genome ‘omics’ including transcriptomics, proteomics wide linkage analysis (Linkage V Panel) and metabolomics are well suited for in which all the autosomes were excluded identification of host-derived biomarkers. and several suggestive linkage peaks The chemical uniformity of RNA makes it were identified on chromosome X. Fine a particularly attractive target, and RNA- mapping refined the chromosome X linkage Seq enables detailed measurement of peak (Zmax = 1.7 at zero recombination; transcripts over a large dynamic range. DXS8041) to a 17.2-Mb interval flanked One promising category of biomarkers are by the markers rs2050030 and DXS1227. microRNAs, which exhibit several attractive A disease haplotype segregates with all characteristics: Easy extraction from blood affected males and obligate female carriers. or serum, reduced transcriptional complexity Interestingly, this region overlaps both the

63 2015 AGTA Conference

CMTX3 and CMTX4 disease loci. Whole in divergent taxa, such as the central exome sequencing did not identify any bearded dragon (Pogona vitticeps) can mutations in the coding region of genes in provide important insight into the essential the CMTX3 or CMTX4 locus (including the molecular mechanisms and evolution of known CMT gene AIFM1). cytokine regulation. The genome of P. vitticeps was recently sequenced, but We present analysis of this linkage region many immune genes were not annotated. using whole genome sequence (WGS) of an Automated annotation techniques are not affected male to identify the causative muta- sensitive enough to detect rapidly evolving tion. We confirmed the absence of mutations genes, such as cytokines, and they must in the coding regions of genes within the be manually curated. Consequently, key linked region. We examined the region for cytokines have not yet been characterized structural variation (SV) and non-coding vari- in P. vitticeps. In order to characterize ants. Four candidate variants were identified cytokine expression and regulation in that localise within the critical disease inter- the immune response in P. vitticeps, we val. They include an inversion that disrupts first need to determine the complete the CXorf48 gene, a deletion in the 5’UTR nucleotide sequences of key cytokine of the FGF13 gene, a deletion upstream of genes. We sequenced the transcriptomes the SMARCA1 gene, and a non-coding point of unstimulated and stimulated spleen cells mutation in the 3’UTR of the SOX3 gene. and constructed full-length transcripts from No novel candidate mutations or SV were short paired-end read RNA sequencing identified that affect the AIFM1 gene in the data using the Trinity platform. De novo CMTX4 locus. The variants identified in this transcript reconstruction and conserved family may help to identify the unknown gene synteny analysis allowed us to identify affected in CMTX3 or be a new disease candidate cytokines that are missing locus on chromosome X. Future studies will from the P. vitticeps genome annotation, confirm if these candidate variants are a including interleukin 2 (IL2), for which pathogenic cause of axonal CMT. transcriptional activation events have been well documented in mouse and human. POSTER 5 Identification of key cytokines in P. vitticeps DE NOVO ASSEMBLY AND will reveal cytokine genes conserved across IDENTIFICATION OF CYTOKINE GENES vertebrates and thus likely to be essential to IN POGONA VITTICEPS the vertebrate immune response. In addition, the nucleotide sequences of key cytokines Alexandra Livernois, Alexie Papanikolaou, in P. vitticeps will enable us to characterize Kristine Hardy, Renae Domoaschenz, Sudha the immune response for comparison with Rao, Tariq Ezaz, Arthur Georges, Stephen mammals. Sarre, Janine Deakin

ABSTRACT Cytokines are small proteins that play an important role in the immune response in vertebrates. Extensive study of cytokine genes in mammals has demonstrated the importance of their time and place of expression to achieve an appropriate immune response. Similar analyses

64 2015 AGTA Conference

POSTER 6 Conclusion: This study identified the novel long non-coding RNA RP1X as an N-Myc THE NOVEL LONG NON-CODING RNA target gene. RP1X up-regulates DEPDC1B RP1X PROMOTES NEUROBLASTOMA gene expression, leading to N-Myc protein CELL PROLIFERATION BY STABILIZING phosphorylation at Serine 62, N-Myc N-MYC PROTEIN protein stabilization and neuroblastoma cell proliferation. Bernard Atmadibrata, Pei Yan Liu, E-Young Tan, Jesper Maarg, Andrew Tee, POSTER 7 Jo Vandesompele, Pieter Mestdagh, Marcel Dinger and Tao Liu DNA METHYLATION IN PATIENTS WITH A STRONG FAMILY HISTORY OF EARLY ABSTRACT ONSET CORONARY HEART DISEASE Rationale: MYC oncogenes encode transcription factors commonly Vicky Cameron, AP Pilbrow, AF Faatoese, overexpressed in many types of cancers. RW Troughton, AM Richards and J.F Neuroblastoma is a childhood solid tumour Pearson that arises from neural crest cells of the sympathetic nervous system. MYCN gene ABSTRACT amplification, which leads to overexpression Coronary heart disease (CHD) is a leading of N-Myc protein, occurs in approximately cause of morbidity and mortality in Aoteoroa/ 40% of patients with aggressive New Zealand. Having a first-degree relative neuroblastomas. with early-onset CHD doubles the risk of CHD [1]. The mechanism by which our Results: RNA-sequencing identified 5 environment influences the expression of transcripts, including RP1X, considerably our genes is through epigenetics, such as differentially expressed between MYCN methylation of the DNA. DNA methylation gene amplified and non-amplified human has been associated with CHD at specific neuroblastoma cell lines. Knocking-down loci [2, 3], but there have ben no previous N-Myc expression with siRNAs reduced studies of DNA methylation in a cohort with RP1X gene expression, and chromatin a very strong family history of early onset immunoprecipitation assays showed that CHD. The Christchurch Family Heart Study N-Myc protein bound to the RP1X gene (FHS) patients are unrelated individuals promoter. Affymetrix microarray studies who have had a documented premature revealed that DEPDC1B was one of few CHD event (<50 years in men, <60 years genes considerably down-regulated in in women) and who have at least one neuroblastoma cells after transfection first-degree relative with an early CHD with RP1X siRNAs. Depletion of RP1X or event (total 80 to date). Controls from the DEPDC1B significantly reduced N-Myc Canterbury Healthy Volunteers (HVOLs) protein phosphorylation at Serine 62, N-Myc with no prior diagnosis of any cardiovascular protein stabilization and neuroblastoma disease were age and gender-matched to cell proliferation/survival. In human the FHS patients. neuroblastoma tissues from the European Neuroblastoma Research Consortium, high To assess the association of DNA levels of RP1X gene expression correlated methylation on risk for those in this tightly with N-Myc target geneset activation, defined cohort with a very strong family DEPDC1B gene expression and poor patient history of early onset CHD, we have prognosis. performed a screening study examining

65 2015 AGTA Conference the methylation status of peripheral by its breakpoint and will finally get a human blood samples from 71 people, 50 from genome inversion map. the FHS and 21 HVOLs, using illumina’s HumanMethylation450 beadchips (485,000 POSTER 9 methylation sites, covering 99% of RefSeq genes). The results of our analysis pipeline WHOLE GENOME VS TARGETED confirm associations of methylation status SEQUENCING: WHICH PROVIDES with age and gender previously reported BETTER DIAGNOSTIC YIELD FOR for population studies, and no differences DISEASE-CAUSING VARIANTS? in methylation associated with white Mark Pinese, Marcel E Dinger and Mark J cell profiles between FHS and HVOLs. Cowley Differential methylation between FHS and HVOLs was observed at 24 sites (adjusting ABSTRACT for age and gender, FDR corrected p<0.05). Many diagnostic laboratories have adopted Using genotypes derived from the Illumina targeted next generation sequencing (eg iSelect Cardio-MetaboChip (220,000 TruSight One, or Whole Exome), and SNPs associated with cardiovascular have made significant advances in terms and metabolic traits), we can show clear of diagnostic yield and turn around time. evidence of sites where methylation is Although Whole Genome Sequencing associated with genetic variations both local (WGS) is recognised for its broad- and and at a distance, known as methylation uniform-depth of sequencing coverage QTLs. across the genome, its diagnostic performance has largely not been POSTER 8 demonstrated. Here, we evaluate the diagnostic performance of WGS compared USING POPULATION SEQUENCING DATA to targeted sequencing in a controlled TO INVESTIGATE THE EVOLUTIONARY clinical diagnostic setting. ROLE AND FUNCTIONAL IMPACT OF INVERSION POLYMORPHISMS We selected two modern best-in-class technologies for comparison: Illumina HiSeq Haojing Shao, Devika Ganesamoorthy and X WGS at 30X depth and the Illumina Lachlan Coin TruSight One target panel at 78X average ABSTRACT target depth. Reads of NA12878 reference Inversions are a class of functional DNA from each technology were processed polymorphisms in human which remains by a common GATK best practices pipeline, poorly understood. In particular, detection and calls compared to the Genome in a of inversions mediated by Non Allelic Bottle consortium gold-standard variants. Homologous Recombination(NAHR) is To simulate the performance on clinically complicated by the fact that they are relevant variants, comparisons were flanked by long (>2000bp) Inverted Repeat restricted to coding exon variants that were sequence and so hard to efficiently detect within 30 bases of a Human Gene Mutation by next generation sequencing method. Database pathogenic locus. We generated a whole genome IR map HiSeq X WGS provided superior diagnostic to cover all possible NAHR inversions. yield to the TruSight One panel across all By target capturing and third generation mutation types, at both homozygous and sequencing 407 IR regions, we could apply heterozygous loci. For single nucleotide computational method to genotype inversion variants, WGS data provided a sensitivity

66 2015 AGTA Conference of 99.9%, with one false positive in over 1.6 independent. In contrast, high-throughput million bases examined; despite its greater sequencing (HTS) based methods used for depth, the TruSight One panel only achieved genetic mapping are generally background a sensitivity of 95.5%, with a higher false independent. Some HTS methods such positive rate. The difference was even more as Genotyping-by-sequencing (GBS) and marked for multi-nucleotide substitutions and RAD-seq use DNA for mapping, while other small insertions/deletions, in which WGS methods such as BSR-seq and MMAPPR detected 19/19 events, whereas the panel use RNA. Current DNA-based methods only identified 13/19 (sensitivity of 68.4%). barcode DNA extracted from each individual in the mapping population to construct the For the detection of variants near known sequencing library, and RNA-based methods disease loci, WGS provides greatly superior construct a separate library from each of diagnostic sensitivity to panel sequencing, two pools, namely mutant and normal. Both despite its lower average depth. This approaches provide high resolution maps to difference is especially marked for multi- identify causal loci, but are not cost-effective nucleotide substitutions and indels, for which for screening a large number of mutant the sensitivity of panel sequencing is only families such as may be recovered from approximately 70%. an enhancer/suppressor screen. Here we In summary, our results anticipate that WGS present a low-resolution, but cost-effective, will improve diagnostic yields over targeted HTS-based method for genetic mapping. panels, even when considering only those For each new mutant we pooled tissue genes targeted by the panel. from phenotyped individuals to create a mutant pool and a normal pool. We adapted POSTER 10 the original GBS method to construct sequencing libraries, prepared libraries for BULKED SEGREGANT - GENOTYPING- several pairs of pools and determined rough BY-SEQUENCING: COST-EFFECTIVE map positions. Our method is cheaper than AND BACKGROUND INDEPENDENT the current GBS protocol, easier than using GENETIC MAPPING OF MUTANTS AND RNA for library construction, and without QTL sampling biases inherent in using RNA expressed in a certain tissue type(s). We are Kokulapalan Wimalanathan, Rebecca currently fine mapping the intervals identified Weeks and Erik Vollbrecht by BS-GBS, and extending the method to ABSTRACT map natural modifiers. Here we present Genetic mapping of new mutants, which the pipeline and results from these genetic allows us to map a mutant phenotype to mapping efforts in maize. a causal locus or loci in the genome, is a crucial step in forward genetics. Construction of a mapping population that consists of mutant and normal individuals is essential for genetic mapping. The mapping population can be used by different high-throughput methods for genetic mapping. Single Nucleotide Polymorphism (SNP) arrays and Sequenome-based methods detect presence and absence of pre-discovered SNPs, and therefore are not background

67 2015 AGTA Conference

POSTER 11 the performance to existing annotations, and creating a designated set of high-confidence MAIZE - GO ANNOTATION METHODS functional annotations for maize genes. EVALUATION AND REVIEW (MAIZE- Review of these annotations by experts GAMER) will substantially improve the confidence of annotations predicted using the pipeline, so Mr Kokulapalan Wimalanathan, Carson we are designing a user-friendly system to Androf and Carolyn Lawrence-Dill leverage crowdsourcing for manual review ABSTRACT of the predicted GO annotations. We plan Maize is an important agricultural crop to expand and adapt the pipeline to other - based on metric tons, maize is the #1 species once it has been implemented and production grain crop in the world (http:// tested in maize. faostat.fao.org). Largely as a result of plummeting sequencing costs, the growing POSTER 12 availability of whole genome sequences GENETIC CHARACTERISATION OF THE has led to an increased demand for high- EVOLUTION OF A NOVEL METABOLIC throughput omics studies. Interpretation of FUNCTION IN YEAST the results of these studies often requires the analysis of functional annotations Åsa Pérez-Bercoff, Tonia L. Russell, Philip of genes. The Gene Ontology (GO) is J.L. Bell, Paul V. Attfield and Richard J. a widely used database that consists of Edwards terms that describe gene product function and property. The majority (~99%) of the ABSTRACT GO annotations in maize are inferred from One of the central questions in biology electronic annotations by high-throughput is how novel biochemical pathways pipelines such as Ensembl. On the contrary, evolve. In sexual populations this is only about half of the GO annotations in particularly complex. Recombination and Arabidopsis are inferred from electronic the random assortment of chromosomes annotations. Clearly there is a need to during meiosis means that an enormous evaluate the confidence of existing GO diversity of genotypes can be generated annotations for maize and improve the from the standing variation present in a overall quality of functional predictions in population. The relative contribution of this this well-studied species. Here we present a genetic variation versus novel mutations pipeline that annotates GO terms to maize in the evolution of novel traits is yet to gene models using multiple functional be established. We are using laboratory- annotation methods along with an evaluation evolved yeasts to investigate the early of confidence for these annotations. stages of evolving a novel metabolic Our pipeline uses three approaches to activity, using a novel approach of mapping assign GO terms: BLAST-based methods, population metagenomic data onto multiple functional domain-based methods, and ancestral genomes. advanced methods (machine learning We are working with a population of and statistical approaches). Using a test Saccharomyces cerevisiae generated dataset that contains high-quality manual from a starting population consisting of 30 annotations from MaizeGDB (~750) and S. cerevisiae strains. The population was reviewed annotations from UniProt (~6500), grown under selection on xylose minimal we are in the process of evaluating the media (XMM), undergoing sexual mating performance of our approaches, comparing

68 2015 AGTA Conference every two months for 1463 days, in which POSTER 13 time it developed the ability to effectively utilise xylose – a trait not present in any AN IMMUNE TRANSCRIPTIONAL of the ancestral strains. We are using a REGULATORY NETWORK APPROACH combination of PacBio long-read single TO MULTIPLE SCLEROSIS molecule real time (SMRT®) and Illumina Margaret Jordan, M. Gresle, L. Laverick, short-read sequencing to investigate the D. Stanley, L. Smith, T. Spelman, H. evolution of this population. As a pilot Butzkueven and A.G. Baxter study, three of the ancestral strains from the starting population, including two ABSTRACT diploids, were sequenced with the new Multiple Sclerosis (MS) is the most common PacBio RSII long read sequencing service disabling neurological disease affecting at the Ramaciotti Centre for Genomics. young adults in developed countries. High quality complete genomes were Approximately 25,000 Australians and as assembled de novo using the hierarchical many as 2.5 million worldwide have MS genome-assembly process (HGAP3) and of these, 48% have profound or severe using only PacBio non-hybrid long-read disability. It is a complex disease associated SMRT sequencing data, and corrected with the effects of multiple genes in using Quiver. In addition, we have shotgun combination with lifestyle and environmental metagenomic Illumina data from one of factors. Among the environmental the early populations exhibiting adaptation associations are the “latitudinal incidence to growth on XMM. We are developing gradient”, adolescent Epstein Barr Virus methods to map these short-read data onto infection, smoking and obesity, whilst the multiple high quality ancestral genomes in strongest genetic association, with an order to estimate the relative contribution odds ratio (OR) of up to 30:1, is an Anglo- of each ancestor’s genetic variation to Celtic ethnicity compared to an East Asian the evolved population, and to identify background. The strongest individual genetic possible sites of recombination. In addition, risk variation is the carriage of the HLA-DR2 metagenomic data is being mapped haplotype, with an OR of 2.5:1. To date, against the official S. cerevisiae reference genome-wide association studies (GWAS) strain S288c to conduct variant calling to have identified 110 disease-associated identify single nucleotide polymorphisms single nucleotide polymorphisms (SNPs) but (SNPs) that are not present in any of our although a few tagging SNPs used in GWAS genomes. These will be compared to the affect transcript splicing, in general the publicly available “100 yeast genomes” GWAS approach tests disease associations and partitioned into natural variation and with individual variants that are unlikely candidates for novel mutations. By doing to be causal. In addition, comparisons of these comparisons we hope to elucidate family-based estimates of heritability with how the population has evolved to acquire simple models of aggregated genetic risk, its novel characteristics, setting the scene for demonstrate that, to date, the cause of a more in-depth study. only ~28% of the heritability of MS has been identified. In essence, the risk factors’ individual associations with MS are so weak that on their own they contribute little meaningful understanding to the disease liability. We thus hypothesised that the complex genetic phenotype is driven by a

69 2015 AGTA Conference co-ordinated expression of transcriptional provide a new perspective on the aetiology regulatory networks. of complex genetic diseases and offer novel therapies for MS. To test this, we generated a gene co- expression network based on pooled POSTER 14 Affymetrix Human Gene 1.0 ST array analyses of magnetic bead sorted “PROFILING THE PROFILERS”: lymphocytes, including B cells, CD4 and EVALUATING CURRENT CD8 T cells, NK cells and monocytes, from BIOINFORMATICS METHODS FOR 67 untreated relapsing/remitting (RR) MS CHARACTERISING THE DIVERSITY OF patients and 102 Healthy Controls (HC) – MICROBIAL COMMUNITIES a total of 712 microarrays. The Affymetrix , Lavinia Gordon, Ken CEL files were normalised using RMA Naga Kasinadhuni McGrath, Rachael McNally background subtraction in Bioconductor and batch effects were removed using ABSTRACT the nonparametric CombatR algorithm. Microbial diversity profiling through amplicon Variability of transcripts across all arrays sequencing is a popular technique for was ranked by standard deviation and the determining the relative proportions of 19,659 most variable were used for network microbial taxonomic groups in a mixed construction. Application of the WGCNA microbial community. Whilst best-practice algorithm in R generated a weighted gene methods for library preparations and co-expression network of 5,762 nodes and sequencing techniques are well established, 198,937 edges assigned to 16 significantly there are a variety of bioinformatics methods co-expressed modules containing between available to analyse the data. Most of the 23 and 1,690 transcripts each. For each current publications on taxa characterization leukocyte population, the strength of utilize tools like the Quantitative Insights differential expression between patients into Microbial Ecology (QIIME), UPARSE, and HC was assessed, by ranking genes mothur, MEtaGenome ANalyzer (MEGAN) by Mann Whitney U test and ANOVA, in and the MetaGenomic Rapid Annotation order to test for each transcript across using System Technology (MG-RAST). the network. Of particular interest was a Along with these sophisticated tools there group of transcripts we named the “Black” are many databases like Greengenes, module, which was strongly down-regulated Silva, Genbank, User-friendly Nordic in monocytes of patients and contained ITS Ectomycorrhiza (UNITE) databases 12 of 22 highly differentially expressed compiling various sequences. (HDE) transcripts (p<10-34, c2 test). The Black module contains 181 transcripts, of To explore the variance generated using which 14 are encoded by genes adjacent these different methods, we analysed to MS-associated single nucleotide amplicon data from an extreme halophilic polymorpohisms (SNP) and seven encode microbial population from a remote and transcription factors (TF). We now aim to isolated pink salt lake (Lake Hiller, Western test our hypothesis by manipulating the Australia) for several targets (Bacterial identified genes in an animal model of MS, 16S:V1-V3, 16S:V3-V4, Fungal ITS1-2, experimental autoimmune encephalomyelitis Eukaryotic 18S, Archaeal V1-V3) across (EAE), as well as by in-vitro techniques, to each analysis pipeline, and compared the control the expression of the whole black microbial profiles generated. module as an integrated unit. This may

70 2015 AGTA Conference

Our results reveal that each method POSTER 16 generates a different outcome, with significantly varied performance at each THE EXPRESSION OF SMALL stage from pair-read assembly, quality NUCLEOLAR RNAS IN HUMAN TISSUES tolerance, and to the accuracy of Taxonomic Hyun Jae Lee, Cas Simons, Joanna Resolution, along with vastly different Crawford, Lachlan Coin, Ryan Taft demands on computational resources. ABSTRACT These comparisons not only highlight the Small nucleolar RNAs (snoRNAs) are a importance of selecting an appropriate class of non-coding RNAs traditionally bioinformatics methodology, but demonstrate assumed to have restricted housekeeping the dangers of comparing microbial profiles role in rRNA and snRNA modifications. More generated in separate studies that have recently, however, various studies have used different analysis tools. uncovered great complexity in snoRNA POSTER 15 biology, warranting a deeper insight into various aspects of snoRNAs. To explore HUMAN WHOLE GENOME SEQUENCING the expression of snoRNAs, we sequenced WITH LOW AND ULTRALOW DNA INPUTS intermediate-sized RNAs obtained from 12 different human tissues, and we showed Kirby Siemering, Jafar S. Jabbari, Lavinia that the expression of snoRNAs could be Gordon, Matt Tinning, A/Prof. Marcel Dinger, determined at much improved resolution, Dr. Maria Lubka-Pathak, Ms Dahlia Saroufim and for the first time, we were able to identify and David Miller the differences in expression of snoRNAs with high sequence similarity, especially ABSTRACT the clusters of snoRNAs playing a key There are several instances in which DNA role in Prader-Willi syndrome. Analysis of availability for next generation sequencing these data sets also revealed that many library preparation is limited. Examples snoRNA loci give rise to various different include: ChIP, FFPE and cell free as well isoforms, including longer RNAs originating as when scarce tissue DNA will be used from SNORD115 cluster and nontemplate for multiple applications. This becomes additions of adenosines to snoRNA 3’ ends. a particular issue for whole genome We further identified novel box C/D and H/ sequencing where good coverage of the ACA snoRNA candidates that have lower whole genome is necessary. To address this expression and less evolutionarily conserved issue, we investigated the characteristics relative to the currently annotated snoRNAs. of libraries prepared from low, ultra-low, Contrary to previous studies, sequencing of and standard fragmented inputs using the small RNAs from 12 different human tissues Rubicon ThruPLEX® DNA-seq kit and revealed that small RNAs derived from sequenced on both the HiSeq 2500 and snoRNAs (sdRNAs) are highly variable in HiSeq X systems. We present data on length. However, we observed that the highly important metrics such as library diversity, abundant sdRNAs with specific length also SNV calls, GC bias and coverage and originate from distinct positions relative to demonstrate the feasibility of sequencing parent snoRNAs. Finally, we show that some previously inaccessible low input samples on sdRNAs with miRNA-like characteristics the HiSeq X system. display tissue-specific expression. These findings suggest that snoRNAs are a dynamic group of non-coding RNAs with

71 2015 AGTA Conference diverse characteristics and physiological POSTER 18 roles. QUANTIFICATION OF CELLULAR POSTER 17 FEATURES USING NOVEL IMAGE ANALYSIS ALGORITHMS EVALUATING THE EFFICIENCY AND REPRODUCIBILITY OF GENOTYPING BY Matloob Khushi and Jonathan W. Arthur SEQUENCING ABSTRACT Rust Turakulov, Jafar Jabbari, Gai Digital microscopes are routinely used McMichael, Paul Gooding in laboratories to understand biological processes. Current technology makes ABSTRACT it easy to automatically generate a very Genotyping by sequencing (GBS) is a large number of images. Researchers sequencing-based method for the large thus require new methods to analyse the scale discovery and genotyping of genetic images precisely and quickly. Manual polymorphisms. It is suitable for population quantification of features in an image is studies, germplasm characterization, subjective and can be prone to human breeding, and trait mapping in diverse error. Therefore, we have developed three organisms. It is amenable to multiplexing novel algorithms to automatically quantify making it high-throughput and cost-effective. various commonly studied cellular features GBS is based on fragmenting genomic using MATLAB (MathWorks USA) language DNA with restriction enzymes, followed and tools. Algorithm-I identifies protein by sequencing a library prepared from locations and their co-localisation stained in the small portion of the genome adjacent different colours: a common technique for to the restriction sites. The AGRF has determining the potential for two proteins to established a multiplex GBS protocol that interact with each other. A graphical interface uses two enzymes, resulting in longer is provided to visualise and fine tune vari- reads and increased sequencing output. ous parameters such as applying median or To complement the protocol, the AGRF has Wiener adaptive filtering to improve signal built an analysis pipeline based on Stacks to noise ra-tio. Command-line mode of the (J. Catchen et al. 2013 Stacks: an analysis script allows batch processing of all images tool set for population genomics. Molecular in a directory. Users can also calculate Ecology. 22:3124-3140). Here we present statistical significance of the observed the results of one study where libraries were protein co-localisations against overlap by made from identical control DNAs sourced ran-dom chance by computing the P-value from E. coli. In that experiment, 48 libraries using the Student’s t-test. Algorithm-II were made from the same DNA source, measures area, length, width, angles, and sequenced and compared. The results shape of stained DNA and the mitotic spindle looked at genotyping errors, batch effects in order to identify the features as-sociated and reproducibility in terms of the numbers with normal or diseased cells. The shape of shared tags. is reported by calculating the eccentricity of the feature, a measure of the degree of circularity or linearity. Algorithm-III measures the length of te-lomeres, as short telomere lengths are known to be associated with a number of diseases. These novel algorithms have enabled us to precisely, efficiently, and

72 2015 AGTA Conference automatically quantify cellular fea-tures for found mutations in cancer patients that large numbers of images. overlap G-quad coordinates in 5’UTRs. We have analyzed the effect of these mutations POSTER 19 on the stability of G-quads using RNAfold software5 and showed that mutations could NON-CODING MUTATIONS IN CANCER either destabilize or stabilize G-quads. We ALTER G-QUADRUPLEX STABILITY AND have validated the effect of some of these AFFECT ITS REGULATORY ROLE IN THE mutations by luciferase reporter assays and 5’ UNTRANSLATED REGION OF MRNA circular dichroism spectroscopy. Among the Mahdi Zeraati, Aaron L. Moye, Mark J. validated mutations is a recurrent mutation Cowley, Jason W. Wong, Tracy M. Bryan, that has been reported in two patients with Daniel U. Christ and Marcel E. Dinger B-cell lymphoma and malignant lymphoma. This mutation destabilizes the G-quad in the ABSTRACT 5’UTR of BCL2; which is an oncogene and Substantial efforts have been made to overexpressed in lymphomas. Luciferase characterize the cancer genome. The advent reporter assays showed that this mutation of next generation sequencing has enabled increases the translation of BCL2. Using a affordable and rapid large-scale sequencing RNA G-quad stabilizing molecule, N-methyl of cancer genomes. The Cancer Genome mesoporphyrin IX (NMM), we showed that Atlas (TCGA)1 and the International the destabilizing effect of this mutation Cancer Genome Consortium (ICGC)2 are is reversible. Our results demonstrate a examples of large-scale cancer genomics potential role of a non-coding mutation in projects to catalogue cancer-associated cancer and show that mutations that effect mutations. Extensive efforts have focused G-quad stability can be targeted, which on characterizing mutations in protein- presents opportunities for highly targeted coding regions, which has improved the therapies. understanding of the pathogenicity of coding mutations. However, knowledge regarding POSTER 20 the effect of non-coding mutations in cancer development is very limited. One possible SMALL RIBOSOMAL SUBUNIT effect of non-coding variations is in affecting PROFILING QUANTIFIES secondary structures in regulatory regions, CONFORMATIONAL CHANGES such as promoter and 5’ untranslated region IN INITIATING AND TERMINATING (5’UTR) of mRNA. RIBOSOMES.

Guanine rich DNA and RNA sequences Stuart K. Archer, Nikolay E. Shirokikh, can fold into a non-canonical four-stranded Thomas Preiss secondary structures called G-quadruplexes ABSTRACT (G-quads)3. RNA G-quads are A large fraction of variation in gene computationally predicted to be prevalent expression in eukaryotes is not accounted in the 3’ and 5’ UTR of mRNAs. This for by differences in mRNA abundance observation strengthens the hypothesis that but rather by regulation of translation these motifs have critical regulatory roles initiation, however we have an incomplete in the posttranscriptional control of gene understanding of this complex process. expression4. Here we employ deep sequencing to profile Using TCGA and ICGC databases, we have nuclease footprints of the small ribosomal subunit, which is the first subunit to make

73 2015 AGTA Conference contact with the mRNA during initiation and driven to a later stage, in that it models compare these to footprints from translating expression as well as gene-interaction by complexes made up of both small and estimating the inverse covariance matrix. large ribosomal subunits in living cells. We Our approach therefore, integrates data to observe stark differences in footprint size for achieve a richer data structure, and utilises small subunits before and after encountering expert data (KEGGpathways) for validation. the start codon, reflecting conformational changes of both mRNA exit and mRNA entry The high dimensionality of the problem has channels. This enabled us to quantify the hampered its application in the past. We relative abundances of initiation complexes overcome this limitation by assuming that at different stages of translation initiation, the inverse covariance matrix is sparse, in vivo. Small subunit footprints were also and divide the problem into two parts; a) identified on the termination codons of estimating the pattern of non-zero entries, ORFs, allowing us to investigate post- and; b) estimating the entries given this termination events leading to ribosome template [2]. This produces a resulting recycling. These results confirm and expand data structure containing genes that are on current structural models of translation differentially expressed or differentially initiation, and provide a framework to probe connected, as well as housekeeping both generalities and transcript-specific genes that are observed to be active in the variations in the translation initiation process. analysed data but are not related to the experimental condition. POSTER 21 We applied this methodology to identify CANCER GENE-INTERACTION ANALYSIS the abnormal gene-interactions in tumor USING A DATA-DRIVEN INVERSE samples containing a point mutation in the COVARIANCE MATRIX APPROACH K-Ras gene (n=65 compared to the wild type n=75). When applying traditional methods, Denis C. Bauer, Robert Dunne, Aidan OLFM4 is identified as differentially O’Brien, William Wilson and Peter Molloy expressed in colon cancer studies, however, our method identifies 133 genes

ABSTRACT in the olfactory transduction pathway to be Gene expression is tightly coordinated for differentially interacting. Separate to their normal biological processes such as cell canonical function, olfactory receptors have division, proliferation and differentiation. been demonstrated to promote cancer cell Misregulation of these processes may invasiveness and metastasis emergence [3]. lead to diseases like cancer. Patterns of This phenotype is also consistent with the gene expression can be studied by way disruption of the Ras protein, which plays of differential gene expression analysis an important role in the propagation of early or regulatory gene network comparisons. signals from the cell membrane into the Recognizing the need for condition- nucleus, as well as the 68 other genes found specific regulatory networks, most recent to be differentially interacting in the cytokine- approaches integrate expression data with cytokine receptor interaction pathway. known protein-protein interaction networks Our method also identifies differentially [1], thereby joining data driven analysis interacting genes in known cancer pathways and databases of accumulated biological such as PI3K-Akt and MAPK. information.

Comparing these approaches, ours is data

74 2015 AGTA Conference

POSTER 22 POSTER 23 FUNCTIONALLY CHARACTERIZING PROMISCUOUS DNA-BINDING OF EVERY SINGLE MUTATION IN YOUR A MUTANT ZINC FINGER PROTEIN FAVORITE GENE CORRUPTS THE TRANSCRIPTOME AND DERAILS ERYTHROID DIFFERENTIATION Alan F Rubin, Stanley Fields, Terence P Speed, Douglas M Fowler Andrew C. Perkins, Kevin R. Gillinder, Melissa Ilsley, Danitza Nébo, Ravi ABSTRACT Sachidanandam, Mathieu Lajoie, Graham W. Although high-throughput DNA sequencing Magor, Michael R. Tallack, Timothy Bailey, has rapidly expanded catalogues of normal Michael J. Landsberg, Joel Mackay, Joel and disease-associated variation, the H. Graber, Luanne L. Peters and James J. functional consequences of most mutations Bieker are unknown. In deep mutational scanning, selection for protein function applied to ABSTRACT a library of protein variants is combined The rules of engagement between zinc with high- throughput DNA sequencing, finger transcription factors and DNA have allowing direct measurement of the activity been partly defined by in vitro DNA-binding of hundreds of thousands of variants of the and structural studies, but less is known protein easily and cheaply. This approach about how these rules apply in vivo. Here helps to bridge the gap between variant we demonstrate how a missense mutation identification and interpretation, enabling in the second zinc finger of Krüppel-like researchers to elucidate sequence-function factor-1 (KLF1-E339D) leads to degenerate relationships at high resolution. The resulting DNA-binding specificity in vivo, resulting in data can be used in a variety of contexts, ectopic transcription and anemia in the Nan from aiding the assessment of clinical mouse model. We employed ChIP-seq and variants to guiding protein engineering. 4sU-RNA-seq (Figure) to identify the direct Despite the growing popularity of deep transcriptional consequences of aberrant mutational scanning, there are few formal DNA-binding events genome wide. We statistical methods available to help analyze expressed wild type and mutant KLF1 zinc these complex datasets. Here we present a fingers in bacteria as GST fusion proteins, novel method for assigning functional scores purified these and used EMSA and surface and statistical significance to all variants plasmon residence to determine the affinities in a deep mutational scanning dataset for various DNA-binding motifs in vitro. based on weighted regression. These We have discovered a novel biochemical methods are implemented as part of Enrich mechanism for disease which also has 2, a user-friendly software package that implications for zinc finger nuclease design makes the initial data analysis accessible to and off target effects. experimental biologists while providing an extensible framework for bioinformaticians manipulating these large datasets. For illustration, we show the results of applying this method to deep mutational scans of diverse targets, including BRCA1 and the WW protein-binding domain of YAP65.

75 2015 AGTA Conference

POSTER 24 wildtype and D45 homozygous offspring of the same age. Analysis of Klf1H350R/- CHARACTERISATION OF NOVEL by flow cytometry showed an increase in HYPOMORPHIC AND NULL MUTATIONS circulating immature red blood cells. In the IN KLF1 DERIVED FROM A GENETIC bone marrow, a lack of mature red blood SCREEN FOR MODIFIERS OF Α-GLOBIN cells was observed. Flow cytometric analysis TRANSGENE VARIEGATION of the spleen from Klf1H350R/- animals revealed an expansion of erythroid cells Andrew C. Perkins, Stephen Huang, and a relative reduction in B and T Cells. Anabel Sorolla, Michael R. Tallack, Harald Electromobility gel shifts assays (EMSA) of Oey, Sarah K Harten, Lucia Clemens- a recombinant Klf1 zinc finger protein with Daxinger, Graham W. Magor, Alex N the H350R mutation showed normal binding Combes, Melissa Ilsley, Kevin Gillinder and to the b-globin promoter sequence but weak Emma Whitelaw binding to the Alas2 intronic enhancer site. ABSTRACT Furthermore b-globin gene expression was Position-effect variegation of transgene near normal whereas expression of other expression is sensitive to the chromatin known Klf1 target genes was decreased. We state. We previously reported a forward will discuss how H350R disrupts function genetic screen in mice carrying a variegated from ChIP-seq and RNA-seq in primary fetal a-globin GFP transgene to find novel liver tissue. Previous studies of the linkers in genes encoding epigenetic regulators. We C2H2 zinc finger transcription factors have named the phenovariant strains Mommes revealed their necessity as structural and for Modifiers of murine metastable epi- regulatory components for the C2H2 class alleles. Here we report positional cloning of transcription factors. Our results show of mutations in two Momme strains which that the second linker of Klf1 has a role in result in suppression of variegation; maintaining the integrity of Klf1 function (at a i.e. an increased percentage of GFP+ subset of Klf1-occupied sites,) and does not circulating red blood cells. Both strains act just as a spacer for the zinc fingers. harbour point mutations in the erythroid specific transcription factor, Klf1. One (D11) generates a stop codon in the zinc finger domain. D11 homozygous mice die in utero of anaemia at 14.5DPC. The other (D45) generates an amino acid transversion (H350R) within a conserved linker between zinc fingers two and three. Homozygous MommeD45 mice have mild compensated microcytic anaemia which models the phenotype in a recently described human family. Mice Carrying the H350R mutation were interbred with Klf1+/- mice. Klf1H350R/- mice have severe perinatal haemolytic anaemia and marked splenomegaly. Furthermore blood haemoglobin content, haematocrit and red blood cell size (MCV) were significantly reduced in Klf1H350R/- mice compared to

76 2015 AGTA Conference

of anterior streak derivatives, such as TUESDAY Sox17, Cer1 and Foxa2, along with a corresponding decrease in expression of 13 October 2015 posterior streak derivatives, such as Mixl1, Mesp1 and Hoxa1/2. EVX1 is hypothesized POSTER 25 to function primarily as a repressor, but the direct targets are unknown. We have ELUCIDATING THE FUNCTION OF successfully used CRISPR-Cas9 to V5 tag THE EVX1 PROTEIN CODING/LNCRNA the endogenous EVX1 protein to perform SENSE-ANTISENSE LOCUS DURING ChIP-seq. We are currently in the process of GASTRULATION optimizing the ChIPseq protocol for the V5 antibody and will report on our progress. Andrew C. Perkins, Charles Bell, Kevin Gillinder, Graham Magor, Lorena di Lisio, To determine whether Evx1as has a function Seth Cheetham, Pierre Tangermann, Paulo beyond that of Evx1, we used CRISPR- Amaral3, Anton Karlsbeek, Michael Tallack, Cas9 to remove the shared promoter region John Mattick and Marcel Dinger of Evx1 and Evx1as. Although removal of the promoter almost completely ablated ABSTRACT Evx1 expression, it resulted in only 50% A number of recent studies demonstrate knockdown of the Evx1as transcript. We the bidirectional transcription is a pervasive subsequently removed the shared promoter phenomenon at mammalian promoters. and the majority of Evx1as transcript (2.3kb). In particular, long non-coding RNAs are By qPCR we were unable to observe a commonly transcribed antisense to highly phenotype in the Evx1/Evx1as KO beyond transcribed protein coding genes. Whether the phenotype observed in the Evx1 KO. or not these lncRNAs have a function other Therefore Evx1as does not function in than regulation of the associated protein trans. This does not however rule out the coding gene remains a topic of debate. possibility that Evx1as may function in cis At the Evx1 locus, the homeodomain to regulate Evx1. However, we believe that protein EVX1 and the lncRNA, Evx1as, are Evx1as most likely represents functionally spatially and temporally co-expressed from inconsequential transcription from a a bidirectional promoter during gastrulation. bidirectional promoter. Homologs of the transcription factor EVX1 have previously been shown to regulate POSTER 26 anterior-posterior patterning, however the direct and indirect targets of EVX1 are TWO ORTHOGONAL PROCESSES unknown. UNDERLIE SURVIVAL OF PATIENTS WITH RESECTED PANCREATIC CANCER: To elucidate the potential function of EVX1 A SUCCESSFUL APPLICATION OF and Evx1as during mouse gastrulation, EXPRESSION DECONVOLUTION we performed extensive CRISPR-Cas9 mediated manipulation of the Evx1 locus Mark Pinese, Mark J Cowley, and Andrew V in mouse embryonic stem cells (mESCs). Biankin, for the Australian Pancreatic Cancer As hypothesized, loss-of-function gene Genome Initiative editing of Evx1 resulted in an anterior- posterior patterning during gastrulation ABSTRACT defect as determined by mRNA-seq. EVX1 Identifying the biological processes that KO resulted in an increase in expression determine cancer patient survival remains a

77 2015 AGTA Conference substantial challenge, particularly in highly POSTER 27 heterogeneous tumours like pancreatic ductal adenocarcinoma. Gene expression Discovering genomic hotspots measurements are an attractive starting for columnar growth in Malus x point when searching for processes active in domestica with GBS a cell, as they integrate diverse information Cecilia H. Deng, Wirsich M, Brauksiepe on genetic and epigenetic status, metabolic B, Hilario E, Schröder M-B, Chagné D and flux, and signalling activity. However, this Braun P mixing of inputs makes it very challenging to deconvolve a transcriptional state into ABSTRACT its component biological processes, and Columnar fruit trees, also known as urban provide a simple biological interpretation of a fruit trees, are bred to grow upright in a complex transcriptional mix. narrow column. These types of trees bear fruit on the main stem with short spurs, Methods for expression deconvolution based rather than growing long side branches. on the non-negative matrix factorization Their elegant shape and space-saving (NMF) have recently achieved popularity, growth pattern make them suitable for but are almost universally used in modes landscaping in small gardens or patios in that induce artificial clustering in the urban areas with limited space. The first data, and so may not reflect the true series of columnar apples were released underlying expression dynamics. There is in 1989 (the “Ballerina” series), but further little guidance on how to best apply NMF breeding activities have resulted in better techniques for expression deconvolution, quality cultivars. Previous research shows particularly when searching for a region on the Malus chromosome 10 (Co transcriptional programs that are associated region) plays an important role in generating with a covariate, such as patient survival. the columnar tree phenotype compared We describe a successful application of to normal growth apple trees, due to the NMF to identify two orthogonal signatures insertion of the Ty3/Gypsy retrotransposon of survival in pancreatic cancer, that “Gypsy-44” into the transposon “Gypsy-33”. correlate closely with the discrete processes Our GWAS (genome-wide association of proliferation and the epithelial-to- study) experiment with GBS (genotyping- mesenchymal transition. Both signatures by-sequencing) of more than 200 apple validate in independent cohorts, and are trees not only confirmed a high association also prognostic in other solid cancers. between the columnar phenotype and The novelty of our approach is its use of a the genomic regions on chromosome sparsity constraint on the NMF, that does 10, but also revealed a genomic hotspot not promote artificial sample clusters, on chromosome 12. QTL (quantitative and produces decompositions that are trait locus) mapping and genetic map more reflective of smooth activation of construction will improve our understanding transcriptional programs. The approach of the genomic nature of the columnar we present can form a general template for growth in apple. Comparative genomics of expression deconvolution that is particularly the regions of interest may also help our valuable when small, biologically-simple research of columnar growth in other fruit expression programs that are associated crops in the Rosaceae family, such as pear with a covariate are desired. and peach.

78 2015 AGTA Conference

POSTER 28 function. For example, the top three terms are the cellular compartment terms GENE ONTOLOGY ENRICHMENT ‘intracellular’, ‘cytoplasm’, ‘mitochondrion’ ANALYSIS FOR GENE SUBSETS OF and are hence not capturing the regulatory DISTINCTIVE FUNCTION difference of the mitochondrial genome between the different fat types. To focus Denis C. Bauer, Firoz Anwar, William on mitochondrial function, we limit the Wilson, Jason Ross, Peter Molloy background to only cover MitoCarta ABSTRACT genes (1041), however this results in no Gene Ontology (GO) enrichment analysis significant terms being identified. We know can provide insight into the underlying that mitochondria play a crucial role in biology function common to a list of adipogenesis and hence the two tissues interesting genes, e.g. differentially are likely to have different function. This expressed (DE) genes. To determine highlights the need for constructing a whether a GO term is over/under- purpose-built background. representated in this set, one needs to Starting from the GO terms associated with compare the frequency of this term in a set MitoCarta genes, we include all genes that of genes considered baseline (background). share these GO terms as the background, The choice and size of the background which reveals 163 enriched terms. Extending can influence the identified terms and this approach, we also remove the GO their significance, potentially resulting in terms not associated with MitoCarta genes misleading biological interpretation. This as they are not able to reach significance problem is specifically pronounced in and impede the multiple testing correction. situations where enrichment is performed This final approach of tailored background in an already very distinctive subset of the and pruned GO tree results in 226 enriched. genome, e.g. mitochondrial proteins. These new terms where particularly enriched We therefore propose a new approach in GO categories containing less than 500 for selecting the background and pruning genes. These considerably more specific the GO term tree for the analysis of gene GO terms relate to relevant biological subsets with distinctive function. To compare processes such as beta-oxidation of fatty our method against traditional approaches, acids and acetyl-CoA metabolism. This we perform GO enrichment analysis using finding suggests that by constructing a goseq [1] on 93 differentially expressed purpose-built background set of genes and a (DE) mitochondrial genes (mitochondrial pruned GO tree we are able to identify some protein compendium, MitoCarta [2]) between fundamental difference in the catabolism visceral and subcutaneous adipocyte cells of fats between visceral and subcutaneous in human (3 matched biological replicates) adipocytes. using different background and pruning approaches.

Performing the GO analysis on the 93 DE genes using the full set of GO terms and all non differentially expressed genes in the genome as background (57,818) we obtain 161 significant terms. However, the most significant terms cover broad mitochondrial

79 2015 AGTA Conference

POSTER 29 ADAR3-deficient compared to control mice; however there were pronounced differences THE EFFECT OF ADAR3 DEFICIENCY ON in the editing frequencies of individual sites. RNA EDITING IN MOUSE HIPPOCAMPUS 16 (5.7%) of the candidate editing sites Dessislava Mladenova, Lotta Avesson, Guy showed statistically significant differences Barry, Mark Pinese, Bryce Vissel and John in editing level between the Adar3 wild type S. Mattick and knockout mice, most of which were located in the UTRs of protein-coding genes ABSTRACT in repetitive retroelement sequences known RNA editing refers to the deamination as Short Interspersed Nuclear Elements of adenosines or cytosines to alter the (SINEs). The functional consequence of sequence of RNA, presumably in response RNA editing in UTRs is not known, although to environmental cues, and is most active reports have suggested effects on transcript in the brain. RNA editing may have played stability and translation. Nine of these an important role in cognitive evolution, as sites (56%) were within predicted miRNA it has expanded greatly in mammalian and seed binding sites. Six of the sites showing especially primate lineages. Adenosine significant changes in editing are located in to inosine (A-to-I) editing is catalyzed by genes implicated in neurological disorders. three members of adenosine deaminase ADAR3 deficiency resulted in a significant acting on RNA (ADAR) protein family. reduction of the editing level of 13 out of the Dysregulation of A-to-I editing has been 16 sites, indicating that ADAR3 can enhance associated with a number of neurological the editing frequency of specific sites in vivo. and neurodegenerative disorders and polymorphisms of the vertebrate-specific In conclusion this is the first report indicating ADAR3 are strongly associated with extreme that ADAR3 can modulate the RNA editing old age. levels of specific genes in vivo.

ADAR3 catalytic deaminase activity has POSTER 30 not been detected in vitro, and ADAR3 is thought to act by blocking the action of HARNESSING THE POWER OF SERUM ADAR1 and/or ADAR2. Here we asked how MICRORNAS TO PREDICT SURGICAL ADAR3 deficiency affects RNA editing in OUTCOME FOR WOMEN WITH OVARIAN the mouse hippocampus. We generated CANCER Adar3 knockout mice and performed RNA sequencing from mouse hippocampal tissue. Jaynish S Shah, Gregory B Gard, Jean Yee Yang, Patsy Soon4 and Deborah J Marsh1 On average 49.8 million reads per mouse were obtained. 800 A-to-I(G) sites met the ABSTRACT criteria of sufficient coverage (>20 reads) Introduction: Optimal cytoreduction, and an average editing level of at least 5% complete surgical rem¬oval of the tumour, per sample. Additional stringent criteria is one of the most important prognostic were applied to obtain 282 high confidence factors for women with ovarian cancer (OC), candidate editing sites, the majority of which yet remains challenging to predict prior to (61.7%) occurred in untranslated regions surgery. A molecular tool to assist in this of mRNAs (UTRs), while only 13 sites were prediction, in combination with other clinical located in protein-coding sequences. There factors, could help plan surgery and decide was no global change in the editing level of the best treatment options for individual patients. Circulating short RNAs, called

80 2015 AGTA Conference microRNAs, have been referred to as a Both microRNAs individually outperformed ‘gold mine’ of non-invasive biomarkers due CA-125 (AUC: 0.73 or 0.77 vs 0.69) in to ease of access, remarkable stability and predicting surgical outcome. The addition of disease-specific expression. CA-125 to both microRNAs increased the AUC to 0.81 with classification accuracy of Aims: (1) To test if circulating microRNAs 77%. In conclusion, our data suggest that could separate women with OC from healthy a combination of microRNAs and CA-125 women and, further, predict their surgical could predict surgical outcome for women outcome. (2) To compare the performance with OC. of CA-125, a serum protein used for the diagnosis of OC, with candidate microRNAs POSTER 31 for both purposes. APPLICATION OF GENOME WIDE Materials and Methods: The study was ANALYSIS IN DEVELOPMENTAL EYE conducted in training and test sets DISEASE containing 15 healthy, 15 optimally and 13 suboptimally cytoreduced (>10 mm) Ivan Prokudin, Anson Cheng, Rebecca OC patients each. Pre-surgical sera Greenlees and Robyn V. Jamieson were sourced from the Kolling’s Tumour

Bank. Expression of 167 microRNAs was ABSTRACT Developmental eye diseases cause measured using the Serum/Plasma focus severe visual abnormalities and include panel (Exiqon) in the training set. A total microphthalmia, coloboma, anophthalmia, of 48 microRNAs including promising anterior segment dysgenesis and cataracts. candidates, endogenous reference A number of disease genes including microRNAs and various controls were SOX2, OTX2, CHX10, BMP4 and RAX, are validated using the Custom Pick-&-Mix known to contribute to these conditions, panel (Exiqon) in the test set. Levels of the whereas the majority of disease genes serum biomarker CA-125 were measured are yet to be revealed. Identification of by sandwich ELISA (R&D Systems). Data causative genetic factors is often hampered were analysed using GenEx software (v 6.0, by variable penetrance and expression of Exiqon) and the statistical language R using these conditions which may be linked to ‘limma’, ‘ROCR’ and ‘ClassifyR’ packages. the multigenic nature of the diseases and Results and Conclusions: Predictive features presence of genetic modifiers. and their parameters were estimated from Various signalling pathways including WNT, the training set data using machine learning BMP, Retinoic Acid, Notch, Hedgehog algorithms DLDA and SVM, and their and others are critical for embryonic eye performance was measured on the test set. development. Many different genes play In addition, robust performance measures important roles in these signalling pathways. were calculated from the pooled data using We propose delineation of the components 100-times 4-fold cross validation. Four of these signalling pathways and other microRNAs were differentially expressed in important components in developmental the sera of healthy women compared to OC eye disease in a genome-wide approach patients at 5% false discovery rate (FDR), for interrogation of genetic causes in these but their performance was lower than CA- conditions. 125. Two microRNAs were elevated by more than 1.5-fold in suboptimally compared to We applied whole exome sequencing in optimally cytoreduced patients at 5% FDR. a cohort of over 25 patients displaying

81 2015 AGTA Conference developmental eye disease. Subsequent accessible or samples are from a mixed analyses were combined with data obtained cell population, such as blood. Thus, even from 123 similarly affected patients from if methylation contributes to the disease UK10k project (http://www.uk10k.org). process, the signal may be absent in unaffected tissue or is diluted in a cell Bioinformatic analysis revealed a cluster of mixture and thus difficult to detect. genes enriched in these conditions. Multi- genic inheritance pattern was observed for To understand the effect of methylation at least 2 familial cases. on disease in a complex region such as the MHC, the relationship between the Subsequent functional analyses to genetics, cell-type specific methylation and interrogate the ability of identified mutations methylation differences between individuals to alter expression of the corresponding need to be explored. Using cell-sorted, pathways were performed. Mutations were 450k methylation data from the blood of 6 cloned in and subjected to luciferase assay individuals we show that within the MHC with the corresponding reporter system. there is a subset of CpGs that are highly Our findings demonstrate a key approach in cell-type specific, whilst another subset hypomorphic of novel critical genetic factors varies amongst individuals but is relatively in developmental eye diseases. uniform between cell types, presumably due to different genetic backgrounds between POSTER 32 individuals. A recently published set of MHC CpGs associated with MS falls almost DISENTANGLING GENETICS AND exclusively in the subset of CpGs that vary METHYLATION IN THE MAJOR between individuals in the cell sorted data. HISTOCOMPATIBILITY COMPLEX We have obtained a cohort of individuals with both 450k methylation array data and Jovana Maksimovic, Damjan Vukcevic, imputed HLA alleles to elucidate whether Justine Ellis, Stephen Leslie, Alicia Oshlack methylation differences between individuals ABSTRACT are completely associated with certain Many GWAS of autoimmune diseases HLA haplotypes or whether there is within- have identified SNP associations in the haplotype variability that may moderate major histocompatibility complex (MHC) disease risk. region, particularly within and proximal to the human leukocyte antigen (HLA) genes. Certain HLA alleles have also been shown to be associated with various autoimmune diseases, such as multiple sclerosis (MS). However, many individuals who harbour known risk alleles do not develop autoimmune disease, suggesting that the genetic risk is modulated by other factors.

Epigenetic marks such as methylation might play a role in modifying genetic risk; although, unlike DNA, methylation varies substantially between cell types. This is problematic if the affected tissue is not

82 2015 AGTA Conference

POSTER 33 at diagnosis for endometrial cancer and length of the EIG121 STR short allele when A NOVEL VARIABLE SHORT adjusted for BMI (HR = 1.055, p = 0.019). A TANDEM REPEAT IN THE UPSTREAM longer short allele is correlated with earlier REGULATORY REGION OF THE age at diagnosis for endometrial cancer with ESTROGEN-INDUCED GENE EIG121 IS the median difference in age at diagnosis POTENTIALLY INVOLVED IN CANCER for endometrial cancer being 6 years when RISK length of the STR is dichotomised around 37 repeats (p = 0.0085). No association was Katherine A. Bolton, Elizabeth G. Holliday, identified between the length of this STR Mark McEvoy, John Attia, Anthony Proietto, and breast cancer risk (p = 0.982) or age at Geoffrey Otton, Nikola A. Bowden, Jason P. diagnosis for breast cancer (p = 0.302). Ross, Kelly A. Avery-Kiejda and Rodney J. Scott We have uncovered a highly polymorphic STR in the upstream regulatory region of ABSTRACT EIG121 which appears to be associated with The estrogen-induced gene 121 (EIG121) a lower age at diagnosis for endometrial encodes a transmembrane protein that cancer. This novel variable STR could be has been associated with endometrial, responsible for altered EIG121 expression pancreatic and gastric cancers. In and have implications for disease risk in genome-wide investigations, using the endometrial and other cancers. recently created Short Tandem Repeats in Regulatory Regions Table (STaRRRT)[1], we POSTER 34 found that EIG121 contains a short tandem repeat (STR) located in the upstream DELETERIOUS PASSENGER MUTATIONS regulatory region. Polymorphic STRs can AS A MARKER FOR PROGRESSION alter levels of gene expression affecting TOWARDS LIVER CANCER transcription and hence protein function. This can have major consequences for Magdalena A. Budzinska, Thomas Tu, disease risk and severity. Fabio Luciani and Nicholas A. Shackel

The aim of this study was to analyse ABSTRACT the variability of the EIG121 STR and to Hepatocellular carcinoma (HCC) is determine any association between its associated with hundreds of passenger length and the risk of developing endometrial mutations, which have generally been or breast cancer. ignored but can alter cell survival and thus may act as marker for cancer progression. In this study, DNA from 204 endometrial A previous study has shown that deleterious cancer patients, 206 breast cancer patients passenger mutations (DPMs) occur and 220 healthy controls was analysed. STR frequently and accumulate in tumours. length was determined by PCR, fragment Therefore, we aim to detect DPMs in liver analysis and sequencing. Statistical analysis disease progression and hypothesise included Mann Whitney rank sum tests, Cox that DPMs accumulate in hepatocytes in proportional hazard regression and Kaplan- precancerous conditions leading up to HCC. Meier analysis. We have performed whole exome We found this repeat to be highly variable. sequencing of pre-cancerous liver tissues: Statistical analysis revealed a trend 12 patients with limited level of liver injury towards an association between lower age and 6 HCV-positive patients with liver

83 2015 AGTA Conference cirrhosis. We have also analysed 3 publically POSTER 35 available datasets of paired HCC and surrounding non-tumour liver (total of 148 LOSS-OF-FUNCTION GERMLINE FGFR1 samples). Further, we determined whether MUTATION IDENTIFIED IN A PATIENT these DPMs were likely to alter genes WITH PROLACTINOMA expressed in the hepatocytes by filtering out Mark J McCabe, Anthony R Lam, Tanya J those genes not detected in liver tissue by Thompson, Ann I McCormack and Marcel E microarray analysis. Dinger Increasing numbers of DPMs were observed ABSTRACT in patients with progressively worse liver Background: Familial pituitary tumours disease leading up to HCC (Figure 1). are thought to be rare, occurring in The pattern of observed DPMs in HCC approximately 5% of pituitary tumour is consistent over multiple algorithms cases (Tichomirowa et al. 2011). Germline for scoring deleterious effect, in multiple mutations in MEN1, AIP, p27 and PRKAR1A aetiologies of HCC, and in multiple are known to be involved (Elston et al. datasets. This strongly suggests that DPM 2009), however recently SDHx and GPR101 accumulation is a general mechanism in have been added to the expanding list tumour evolution. Moreover, precancerous of genes implicated in the hereditary alterations were found in non-tumour tissue, predisposition to pituitary tumours (Gill despite being used in prior studies as normal et al. 2014; Trivellin et al. 2014). Utilising paired controls. next generation sequencing technology, In summary, we have shown that DPM we have developed a 300+ gene panel accumulation could act as a potential incorporating genes known to be involved biomarker for risk towards HCC in pituitary tumour pathogenesis, pituitary development without having first to identify embryogenesis and broad cancer genes. rare and unknown HCC driver mutations. We have commenced screening familial pituitary and young sporadic pituitary tumour Figure 1. Increased proportion of DPMs cases with this panel. Using this approach, is associated with HCC compared to we identified a rare missense, heterozygous surrounding non-tumour tissue. variant in fibroblast growth factor receptor 1 (FGFR1) (c.485A>C; p.D162A), in a male with a childhood-onset prolactinoma whose daughter has congenital hypopituitarism. Germline mutations in FGFR1 have been implicated in congenital hypopituitarism.

Aim: To determine whether the identified FGFR1 variant p.D162A is functionally dele- terious using an established culture model, in vitro.

Method: Rat L6-myoblasts which contain very low levels of endogenous FGF re- ceptors and ligands, were transfected with wild-type and mutant FGFR1 pcmv-SPORT6 expression vectors along with a luciferase

84 2015 AGTA Conference reporter driven by 6 tandem repeats of the developments in multiple biomarker-class osteoblast-specific core binding sequences optical barcode counting significantly of the FGF responsive osteocalcin promoter reduce this problem. Recent work from the (Kim et al. 2003). Cells were treated with Weissleder-lab [1] has shown how optical recombinant human FGF2 ligand and then barcode technology can be utilized for lysed for luciferase assay 24 hours later. multiplexed digital counting of proteins, Treatments were conducted in triplicate and and be combined with simultaneous digital cultures repeated three times. counting of nucleic-acids on a single platform. Results: FGFR1 [p.D162A] variant exhibited a 40% reduced function (p<0.001) compared In this study we describe single-molecule to wildtype. digital counting of mRNA, proteins, and gene-fusions in a single simultaneous re- Conclusion: Using our custom pan-cancer action using a few thousand cells as input. gene screening panel, we have identified Cells were fixed, permeabilized, and then a loss-of-function mutation in FGFR1 in a profiled using: (1) PanCancer Pathway patient with a pituitary tumour. Identification Panel (770 mRNAs associated with path- of the same mutation in the daughter and ways linked to key driver mutations) (2) in other families may also implicate FGFR1 PanCancer Immune Profiling Panel (770 in the hereditary predisposition to pituitary mRNAs associated with immune response) tumours. (3) up to 30 key cancer protein targets (+/- phosphorylation) and (4) multiple gene-fu- POSTER 36 sions. The current assay is configured to SIMULTANEOUS “MULTI-OMIC” allow “mixing-and-matching” of different MEASUREMENT OF GENE FUSIONS, biomarker-classes up to a total of 800-plex MRNA AND PROTEINS AT 800-PLEX in a single reaction. All data collected in this USING SINGLE-MOLECULE OPTICAL manner are completely harmonized: within BARCODES the same field-of-view of the detection micro- scope (nCounter system) optical barcodes Michael Rhodes associated with each biomarker class are simultaneously counted, while the software ABSTRACT keeps track of which barcodes originate from Both the ENCODE and TCGA projects DNA, RNA, and Protein. Protein phosphory- showcased the value of quantifying multiple lation-state changes measured by barcodes biomarker classes (DNA, RNA, protein) were also correlated with spatially resolved from cancer tumor samples, providing fluorescent immunohistochemistry images of a much broader view of the underlying the same cells. cancer-biology. Combining multiple data types together into a single correlated analysis, however, is adversely effected by the drastically different methodologies utilized for measurement (i.e., the “detection harmonization” problem). For example, the fluorescence signal intensity obtained from a camera imaging a protein array (e.g., RPPA) is very difficult to correlate directly with an RNA-Seq count of a clonally amplified, cDNA-converted, mRNA molecule. New

85 2015 AGTA Conference

POSTER 37 with the isolation, sequencing and de novo assembly of individual strains of the WILD WINE: METAGENOMIC ANALYSIS dominant wine-associated species also OF MICROBIAL COMMUNITIES DURING being performed in order to aid the analysis. WINE FERMENTATION. Initial results support the view that Peter Sternes, Danna Li, Darek R. Kutyna, uninoculated ferments begin with a diverse and Anthony R. Borneman ecosystem of fungal species, but converge on the wine yeast S. cerevisiae as the ABSTRACT ferment progresses. Notable differences Wine is a complex beverage that is between regions, vineyards and wineries comprised of thousands of metabolites that were also apparent and these can be are produced through the action of yeasts broadly defined by the resulting microbial and bacteria in fermenting grape must. To composition of the wild ferments. ensure a robust and reliable fermentation, most commercial wines are now produced POSTER 38 through the inoculation of freshly crushed grapes with large amounts of the major HIGH QUALITY RNA ISOLATION wine yeast Saccharomyces cerevisiae. FROM RAT SPINAL CORD MOTOR However, there is a growing trend towards NEURONS USING LASER CAPTURE the use of classical, uninoculated or “wild” MICRODISSECTION fermentations in which only those yeasts and bacteria that are naturally associated Prachi Mehta and Renée Morris with the vineyard or winery perform the ABSTRACT fermentation. This generally results in a Objective: To use a Laser Capture far more complex progression of non- Microdissection (LCM) system for isolating Saccharomyces fungal species, with S. RNA from motor neuron cell bodies in a rat cerevisiae only becoming dominant much model of spinal cord injury. later in the fermentation process. The varied metabolic contributions of these Background: Spinal cord motor neurons non-Saccharomyces species have been play a central role in bridging the connection shown to impart desirable taste and aroma between the brain and the skeletal muscles. attributes to wild ferments when compared Information regarding the fate of motor to their inoculated counterparts. Accordingly, neurons below a spinal cord injury (SCI) it is generally believed that differences that have lost their supraspinal input is thus in these resident microflora between essential to understand the pathophysiology vineyards and wineries play a key role in of SCI. As motor neurons represent less defining unique regional expression of wine than 10% of the total cell population in characteristics. the spinal cord, the transcriptional profile of motor neurons below a transection In order the map the microflora of cannot be characterized from spinal cord spontaneous fermentation, metagenomic homogenates. A laser capture approach is techniques have been used to monitor thus ideal for analyzing the mRNA profile the progression of fungal species during in these neurons. Here, we describe the a collection of wild fermentations from protocol, optimized in our laboratory, around Australia. Both amplicon-based ITS for identification and efficient capture of phylotyping and shotgun metagenomics enriched populations of motor neurons by were used to assess community structures, LCM system. This protocol preserves the

86 2015 AGTA Conference

RNA integrity (RIN) of the specimens by homogenate served as a positive control. rapidly freezing spinal cord sections, thus avoiding RNA damage by endogenous and Results: The procedure described above exogenous RNases. produced good quality RNA from fresh as well as PFA treated samples, with a RIN Methods: The rats were anesthetized by above 7 from 800-1200 spinal cord motor intraperitoneal (ip) injection of Euthal (~0.7ml neurons for C2-C3 and C4-C5 segments intraperitoneal for 250g rat) and perfused by LCM technique. Rapidly excising with either only 0.1M PBS or 2% PFA & and freezing the spinal cord segments PBS for about 2 minutes. Segments C2- maintained the RNA integrity. Real-time C3/C4-C5 of the cervical spinal cord were or quantitative PCR (qPCR) using RNA dissected out, rinsed for 10 seconds in isolated from small populations of cells can RNase free water and placed in a cryomold be challenging, due to lack of sensitivity and filled with O.C.T.. The mold was then placed reproducibility. In this report, we demonstrate in a shallow tray containing 2-methylbutane that motor neurons isolated by LCM produce pre-cooled with liquid nitrogen to fast amplified cDNA from as little as 5 ng of RNA. freeze the spinal cord segments to avoid RNA degradation. The O.C.T embedded Conclusions: We conclude that LCM is block was stored at -80°C or immediately an ideal system for the identification and sectioned to produce 50 μm sections on isolation of motor neuron cell bodies to RNase-free slides. The tissue sections generate intact and high quality RNA while were then stained with Azure B to identify preserving their morphology. This system the motor neurons. LCM system (PALM is therefore well suited for downstream Microbeam, Carl Zeiss) was subsequently quantitative PCR studies for gene used to identify and capture the stained expression analysis. In the spinal cord, such motor neurons directly in the cap of a preparation from motor neurons by LCM small microfuge tube containing 30µl will provide valuable tools to advance our guanidine thiocyanate lysis buffer. RNA understanding of the molecular mechanisms was isolated using an RNeasy Micro Kit in SCI as well as in many neurologic and (Qiagen) according to the protocol provided neurodegenerative diseases. for LCM tissue. RNA integrity (RIN) was determined with 2μl sample on a capillary POSTER 39 electrophoresis microfluidics chip (Agilent IMPACT OF GADOLINIUM OXIDE Bioanlyser 2100). The entire procedure was NANOPARTICLES ON VIABILITY OF carried out under RNase-free conditions. HUMAN LIVER-DERIVED CELL LINE In addition, to identify and confirm that the collected cells were motor neurons, Saad Alkahtani, Daoud Ali and Saud Alarifi we performed antibody staining with anti- Chat antibody (Millipore AB144p). We ABSTRACT performed an RT-PCR analysis to confirm Gadolinium oxide (Gd2O3) nanoparticles the presence of microdissected transcripts used in multimodal imaging agents, drug as an additional RNA quality control. cDNA carriers and therapeutic agents. The was synthesised using SuperScript III First objective of this study is to investigate Strand Synthesis System (Invitrogen, Life Gd2O3 nanoparticles induced cell Technologies). We examined expression viability and underlying mechanisms of transcripts for GAPDH, a common in human liver-derived (HepG2) cell reference gene. Total RNA from spinal cord line. MTT (3-(4,5-dimethylthiazol-2-yl)- 2,5-diphenyltetrazolium bromide) and

87 2015 AGTA Conference neutral red uptake (NRU) assays were Despite its industrial importance, there applied to assess cell toxicity due to are very few molecular genetic techniques Gd2O3 nanoparticles. To clarify the that can be applied to B. bruxellensis. As underlying mechanisms, we measured a consequence, the understanding of the the intracellular production of reactive biology of B. bruxellensis lags well behind oxygen species using dichlorofluorescein that of other industrial fungal species, diacetate as a fluorochrome. The level particularly the major industrial yeast, of reduced glutathione was declined and Saccharomyces cerevisiae. However, the malondialehyde level was increased due recent falling costs of next-generation DNA to Gd2O3 nanoparticles. A significant dose sequencing have facilitated an acceleration and time dependent increase of DNA strand in the understanding of B. bruxellensis breaks was found after treatment with biology through comparative genomics. Gd2O3 nanoparticles for 48 hour. Increases in ROS level and the caspase-3 activity A potentially important factor in the were also observed. The DNA strand breaks adaptation of B. bruxellensis to fermentative induced by Gd2O3 nanoparticles were ecosystems is its ability to metabolise almost prevented in HepG2 cells pretreated alternative nitrogen sources, such as nitrate, with NAC (10 mM) for 24 hour. Our results that S. cerevisiae is unable to consume. thus indicated that Gd2O3 nanoparticles Interestingly, it is the only fungal species exert cytotoxic effects on HepG2 cells, most that is commonly found in wine that is likely through oxidative stress. known to assimilate nitrate. However, nitrate assimilation is not a global characteristic POSTER 40 of the species as only around half of B. bruxellensis show this nitrate assimilatory GENOMIC INSIGHTS INTO THE phenotype. NITRATE ASSIMILATION POTENTIAL OF BRETTANOMYCES BRUXELLENSIS In order to determine the basis for the ISOLATES variable ability in nitrate assimilatory potential, forty five B. bruxellensis isolates Ryan Zeppel, Chris Curtin and Anthony were phenotyped and their assimilatory Borneman status correlated with whole genome sequence information. Of twenty nitrate-

ABSTRACT positive isolates, eleven were found to be Brettanomyces bruxellensis is a yeast species that is well-adapted to fermentative heterozygous across the nitrate assimilation ecosystems, thanks to characteristics such gene cluster, while nine exhibited four as ethanol accumulation and tolerance, acid different homozygous nitrate gene alleles. tolerance and the assimilation of limited or Four nitrate-negative isolates were found alternative nutrients. It is known to spoil wine to harbour a deletion of all, or part of, the via the production of phenolic metabolites predicted nitrate assimilation gene cluster. that impart off aromas (e.g. 4-ethylphenol, Surprisingly, a large number of nitrate- 4-vinylphenol); however, the species is used negative isolates were found to possess a to add complexity to the flavour profiles of common homozygous complement of the some specialty beer styles. B. bruxellensis also holds potential as a candidate for use gene cluster. However, it appears that at in continuous fermentations for bioethanol least some isolates of this type switch from production, due to its high stress tolerance nitrate-negative to -positive at a rate of and alternative carbon and nitrogen approximately 1 in every 105 cells. Research metabolism. is now directed towards understanding the

88 2015 AGTA Conference genetic mechanism behind this phenotype (rumen), RHBG-ammonia (intestines) and switch. SLC5A5-iodine (abomasum). The gene encoding a novel, poorly characterized POSTER 41 member of the maltase-glucoamylase family (MGAM-like), predicted to play a role in EPITHELIAL, METABOLIC AND the degradation of starch or glycogen, was INNATE IMMUNITY TRANSCRIPTOMIC highly expressed in the small and large SIGNATURES DIFFERENTIATING intestines. THE RUMEN FROM OTHER SHEEP GASTROINTESTINAL TRACT TISSUES Conclusions: The rumen is probably not a modified stomach or colon with a , V. Hutton Oddy, Richard Ruidong Xiang cornified epithelium, but may be a modified Talbot, Alan Archibald, Phil Vercoe, Brian P. oesophagus with some liver-like and other Dalrymple specialized metabolic functions. ABSTRACT Background: Ruminants are very successful POSTER 42 herbivorous mammals, in part due to ROS MEDIATED APOPTOSIS AND specialized forestomachs, the rumen DNA DAMAGE INDUCED BY BARIUM complex, which facilitates the conversion of NANOPARTICLES IN MOUSE feed to soluble nutrients by micro-organisms. EMBRYONIC FIBROBLASTS

Results: Sixteen gene expression clusters Saud Alarifi, Daoud Ali and Saad Alkahtani were identified from 11 tissues covering the sheep gastrointestinal tract (GIT), two ABSTRACT stratified epithelial tissues and controls. Barium nanoparticles are an important The clustering of the rumen, skin and tonsil industrial material and are widely used in was driven by genes from the epidermal polymer and paints. The objective of this differentiation complex, and genes encoding investigation was to measure possible stratified epithelium keratins and innate apoptosis and genotoxic effects of barium immunity proteins. Consistent with its high nanoparticles in mouse embryonic turnover rate the whole GIT showed a fibroblasts (L929) cells. In vitro cytotoxicity marked enrichment of cell cycle process assays were performed for barium genes (P=1.4E-46), relative to liver and nanoparticles in L929 cells. Mild cytotoxicity muscle, with highest expression in the was observed due to barium nanoparticles cecum followed by colon and rumen. The in L929 cells. Results showed that barium expression patterns of several membrane nanoparticles slightly induced thiobarbituric transporters (Chloride, Zinc, nucleosides, acid reactive substances and superoxide amino acids, fatty acids, cholesterol, bile dismutase and reduced the levels of acids and lactate) along the GIT was glutathione in L929 cells. A concomitant very similar in sheep, pig and humans. by the generation of ROS, the loss of In contrast, short chain fatty acid uptake mitochondrial membrane potential, activation and metabolism appeared to be different of caspase-3 were observed in barium between the species and different between nanoparticles treated L929 cells. In addition, the rumen and colon in sheep. The in the single cell gel test we exposed importance of nitrogen and iodine recycling barium nanoparticles (50-300 μg/ml) at two in sheep was highlighted by the highly treatment times 24 and 48 hour. There was preferential expression of SLC14A1-urea dose and time-dependent increase of DNA

89 2015 AGTA Conference damage observed in the single cell gel test. GEMINI by adding powerful family-based Therefore, the obtained results indicate that filters, as well as an interactive gene list barium nanoparticles may exhibit genotoxic curation tool, and extensive integration with effects in cultured L929 cells, due to disparate annotation databases. This allows induction of oxidative stress. the phenotype and inheritance pattern of the condition to be used to power the analysis. POSTER 43 Seave’s simple web interface means that filtering and searching a database of millions SEAVE: A COMPREHENSIVE VARIANT of variants within a family trio based on 10+ FILTRATION PLATFORM FOR CLINICAL criteria is performed in a matter of minutes. GENOMICS Here we demonstrate the capability of Seave Velimir Gayevskiy, Ying Zhu, Tony Roscioli, through the analysis of genomes derived Marcel E Dinger and Mark J Cowley from >40 families with unsolved genetic diseases. ABSTRACT With rapid gains in accessibility and POSTER 44 dramatically increased diagnostic yields, whole genome sequencing (WGS) is DEAR-O: DIFFERENTIAL EXPRESSION poised to become a routine test in the ANALYSIS BASED ON RNA-SEQ DATA - diagnosis of inherited disease. However, ONLINE the interpretation of genomic variants Zong-Hong Zhang, Naomi R. Wray and remains complex and time consuming. WGS Qiongyi Zhao identifies millions of variants per individual that must be annotated using disparate data ABSTRACT sources, interpreted and reported before Differential expression analysis using high- being made available to clinical geneticists throughput RNA sequencing data is widely or researchers. The recent release of applied in transcriptomic studies. Many GEMINI (Paila et al., 2013) saw this process software packages have been developed significantly improved allowing joint-called for the identification of genes that are variant files to be readily converted into differentially expressed between groups. databases that are then interrogated using There continues to be active development SQL statements. GEMINI combines a in updating of current software tools as number of data sources to comprehensively well as an emergence of many new tools. annotate each variant with its impact, quality, The optimal choice of software tool for prevalence in human populations and clinical differential gene expression analysis is significance. However, as GEMINI is a difficult to evaluate as different software command-line tool, and researchers must tools, and different versions of the same construct custom SQL queries, this software tool, perform differently. Comparison studies is impractical for routine use in a clinical could become out-of-date if one software setting. To enable rapid and easy access to tool were upgraded to a new version. In the GEMINI platform, we have built Seave, order to enable researchers to make timely a web-based variant filtration interface, evaluations of the performance of different designed for genome researchers and software tools with different versions, we clinical geneticists, to easily interpret whole- propose a user-friendly web server that exome and whole-genome-scale datasets, allows end-users to perform a comparison easily handling datasets of hundreds of study using different RNA-Seq tools and patients. Seave extends the power of versions. This web server is built based on

90 2015 AGTA Conference our previous comparative study on software multiple sites at autopsy (CASCADE study). tools for RNA-Seq differential expression The patient, aged just 41 at diagnosis of analysis. Software tools currently included in a Stage I ovarian mucinous tumor, had a this web server are various major versions 26-month progression-free interval, including of DESeq, DESeq2, edgeR and Cuffdiff2. normal CA125 and CA19.9 measurements Data currently used on the server are at 21 months. The primary tumor was publicly available mouse and lymphoblastoid mostly borderline in appearance, with only cell lines RNA-Seq data. The web server a small focus of carcinoma. At autopsy, the will be maintained and updated to ensure carcinoma was widespread in the body, new major versions of current tools or new and whole-genome sequencing data was tools are included. It is open to include new obtained from deposits in the omentum, iliac software tools and new benchmarking data lymph node, para-aortic lymph node and sets in the future. Our web server will serve upper diaphragm. the RNA-Seq community and provide timely guidance for researchers to select optimal Whole genome sequencing (Illumina X Ten, software tool for RNA-Seq differential Kinghorn Centre for Clinical Genomics) and expression analysis. single nucleotide variants (SNVs) identified with multiSNV (Josephidou 2015) are being POSTER 45 used to reconstruct the relationship between metastatic sites. We are exploring the ability ANALYSIS OF METASTASIS of the observed mutant allele frequency PROGRESSION IN HIGH GRADE to infer subclonal structure and history of MUCINOUS OVARIAN CANCER BY progression using various unsupervised WHOLE GENOME SEQUENCING OF clustering methods. Preliminary analysis MULTIPLE AUTOPSY SITES suggests a greater diversity of mutations with lower mutated allele frequencies in , Dane Cheasley, Matthew Wakefield the omental metastasis, and increased Katherine Alsop, Mark Shackelton, Heather private or strongly selected mutations in the Thorne, David Bowtell, Martin Kobel, Prue diaphragm, right illiac node and para-aortic Allan, Andrew Stephens, Tom Jobling, lymph node. Subclonal structure between Sumi Ananda, Orla McNally, Rufaro Diana sites appears complex, and inference of the Jaravaza, Ian Campbell, Clare Scott and relationship between sites is sensitive to Kylie Gorringe chosen thresholds and analysis methods. ABSTRACT Mucinous ovarian carcinomas (MOC) are a distinct histological subtype of epithelial ovarian cancer, with late stage and high- grade MOC often resistant to standard chemotherapeutic regimens. Controversy exists over whether high-grade MOC arise in the ovary or represent metastases from distant sites with secondary ovarian involvement.

We present here the first whole-genome sequencing analysis of a high-grade mucinous ovarian carcinoma collected from

91 2015 AGTA Conference

POSTER 46 and our own preliminary analysis showing these tissues are the most transcriptionally FUNCTIONAL CHARACTERISATION OF active for pseudogenes. This will equip HUMAN PSEUDOGENE TRANSCRIPTION us to investigate the function of the huge USING TARGETED RNA SEQUENCING proportion of the genome thought to be non-functional DNA relics. We will Daniel W Thomson and Marcel Dinger concentrate on two putative mechanisms ABSTRACT for pseudogene function; 1) The shuffling Degenerate copies of genes called of pseudogene exons to produce new pseudogenes are ubiquitous in mammalian chimeric RNA isoforms in the evolution of genomes, in humans there are 14467 new genes, 2) Identification of antisense annotated pseudogenes (Gencode transcription and endogenous pseudogene v21). The most abundant subtype, derived siRNAs (endo-siRNAs) and their ‘processed pseudogenes’ arise through role in silencing somatic retrotransposition. retrotransposition of cellular RNAs. Understanding why pseudogenes are so Colloquially known as ‘jumping genes’, profuse in higher eukaryote genomes and retrotransposons have shaped the human how pseudogene transcription contributes genome over time forming repetitive DNA to cell maintenance and cell defense will estimated to contribute to ~70% of the be crucial to understanding how genomes genome. Transcriptomic studies suggest evolve as well as how cancer cells hijack that at least 9% of pseudogenes are actively these mechanisms for their own proliferation. transcribed. This provides evidence towards pseudogene function as non-coding RNA POSTER 47 where protein coding potential in limited, THE PRESENCE OF THE RS17878362 however these estimates are constrained POLYMORPHISM IN BREAST CANCER IS by the limitation of aligning short read ASSOCIATED WITH A LOW ∆40P53:P53 sequences to pseudogenes, which by RATIO AND BETTER OUTCOME. definition have highly homologous genomic repeats. Kelly A Avery-Kiejda, Brianna C Morten, Michelle W Wong-Brown and Rodney J To investigate transcription of human Scott pseudogenes in a way capable of distinguishing between homologous ABSTRACT transcripts, we have appled RNA Breast cancer is the most common cancer CaptureSeq coupled with long-read in women, but surprisingly it has relatively Pacific Biosciences (PacBio) sequencing. low rates of p53 mutations, suggesting CaptureSeq is a powerful genomics other mechanisms are responsible for approach capable of detecting very low p53 inactivation. We have shown that the abundance transcription, we will capture p53 isoform, Δ40p53, is highly expressed pseudogene regions and apply both in breast cancer, where it may contribute Illumina and PacBio sequencing. This to p53 inactivation (Avery-Kiejda et al. novel application of the technology will Carcinogenesis 2014; 35 (3):586–596). give unprecedented depth and accuracy Δ40p53 can be produced by either of analysis to pseudogene transcription. alternative splicing of p53 in intron 2, or We will perform these analyses on human alternative initiation of translation. The testes and brain samples following reports alternative splicing of p53 to produce of active retrotransposition in these regions Δ40p53 is regulated by the formation of

92 2015 AGTA Conference

G-quadruplex (G4) structures in p53 intron POSTER 48 3, from which the nucleotides forming these structures overlap with a common A HIGH DENSITY SNP-BASED GENETIC intronic polymorphism (rs17878362). The LINKAGE MAP OF SWEET POTATO presence of this polymorphism alters P53 ESTABLISHED BY SCAFFOLDING AN splicing to decrease the expression of the IPOMOEA TRIFIDA (H. B. K.) G. DON. DE fully spliced p53 form following ionising NOVO ASSEMBLY radiation in vitro. Hence, the presence of Chenxi Zhou, Muhammad A. Khan and this polymorphism may be an important Lachlan JM Coin mechanism to regulate the ratio of Δ40p53 to full-length p53 in breast cancer. This ABSTRACT study aimed to determine if the rs17878362 Despite being a critical worldwide food polymorphism was associated with altered resource, sweet potato (Ipomoea batatas Δ40p53 and p53 expression and outcome (L.) Lam.) remains a less well studied crop in breast cancer. We sequenced P53 in due to its genetic complexity (2n = 6x = 139 breast cancers and compared this 90). As a result the availability of sweet to the relative mRNA expression of the potato genomic resources is highly limited Δ40p53 and p53 transcripts using real- [1, 2]. Here we propose a novel method for time qRT-PCR. We found that the ratio constructing high density genetic linkage of Δ40p53:p53 was significantly lower in maps and scaffolding de-novo assemblies tumours that were homozygous for the of polyploid crops using Genotyping-by- polymorphic allele compared to those Sequencing (GBS) data from a large full-sib who were wild-type. Furthermore, there family. By inferring transmission of parental was a lower proportion of polymorphic haplotypes to each progeny, we are able alleles in breast cancers from patients to calculate the recombination fraction who subsequently developed metastasis (RF) between every pair of contigs, from compared to those who did not. Finally, we which we can iteratively cluster contigs show that patients whose tumours were into chromosomes, and finally resolve the homozygous for the polymorphic allele had contig ordering by use of a genetic algorithm significantly increased disease-free survival. which minimizes the RF sum along each These results show that the rs17878362 chromosome. Our method also uses the polymorphism is associated with low shared haplotype structure in the family Δ40p53:p53 ratio in clinical breast cancer together with the allelic depth information to specimens and that this is associated with call high quality SNP genotypes. better disease outcome. Application of our approach to simulated datasets indicates that we are able to accurately reconstruct the correct clustering and ordering of contigs. We have used this approach to construct a high density genetic linkage map consists of 6,428 SNPs from 25 chromosomes by scaffolding a de novo assembly of Ipomoea trifida (H. B. K.) G. Don., the most likely ancestor of sweet potato but of much less genetic complexity (2n = 2x = 30) calls. We anticipate this method may help in constructing high quality genetic linkage maps for more plants.

93 2015 AGTA Conference

POSTER 49 POSTER 50 A NOVEL TECHNOLOGY FOR WHOLE A SINGLE-TUBE NGS LIBRARY PREP GENOME AMPLIFICATION FROM SINGLE WORKFLOW INTEGRATING ENZYMATIC CELLS AND LIMITED MATERIAL FRAGMENTATION RESULTS IN HIGH YIELDS AND LOW SEQUENCE BIAS Martha Hinton (presenter only), Ángel J. Picher, Patricia Garrido, Armin Schneider, Martha Hinton (presenter only), B. Miller, Oliver Wafzig and Luis Blanco M. Appel, .V. Van Kets, B van Rooyen, H. Whitehorn, M Ranik, P. Jones, A. Geldart, R. ABSTRACT Kasinskas and E. van der Walt TruePrime™ is a novel technology dedicated to the amplification of whole ABSTRACT genomes and DNA from various sources. Continuous improvements to library TruePrime is based on the combination preparation for next-generation sequencing of the highly processive Phi29 DNA (NGS) are necessary to achieve the highest polymerase with the recently discovered data quality. One of the crucial steps primase/polymerase TthPrimPol. In this within library preparation is the initial DNA setup, TthPrimPol synthesizes the DNA fragmentation, which can be accomplished primers needed for Phi29 DNApol, which through either mechanical or enzymatic allows for the exponential amplification of processes. Mechanical methods for DNA target DNA. TthPrimPol is a thermostable fragmentation are difficult to scale or member of a recently discovered family automate, and require large investments of enzymes named PrimPol. TthPrimPol in expensive instrumentation. Current is a monomeric enzyme (34 kDa) that enzymatic solutions for DNA fragmentation displays a potent primase activity, preferring typically exhibit sequence bias, provide poor dNTPs as substrates unlike conventional control over fragment length distribution, and primases. This DNA primase activity can are highly sensitive to input amount. be activated by magnesium or manganese ions, having a wide sequence specificity for To address these challenges, we have template recognition. Key advantages of developed the KAPA HyperPlus Library the TruePrime technology include complete Preparation Kit by integrating an enzymatic absence of primer artefacts, insensitivity DNA fragmentation technology with fast and to external DNA contaminations, reduced efficient library construction to provide a amplification bias compared to methods streamlined, easy-to-automate, single-tube using random synthetic primers, and an solution for preparing NGS libraries from 1 exquisite reproducibility when amplifying ng – 1 μg of dsDNA. from single cells or minute DNA amounts. Moreover, TruePrime™ shows superior POSTER 51 sensitivity, is easy to use and works perfectly COOKING UP A PATHWAY ANALYSIS well with commonly used NGS platforms WITH ROAST AND FRY such as Illumina or IonTorrent. We believe that TruePrime will advance human genetic Goknur Giner and Gordon K. Smyth analyses from single cells or otherwise limited input material. ABSTRACT Gene set tests are often used in differential expression analyses to explore the behaviour of a group of related genes. This is useful for

94 2015 AGTA Conference identifying large-scale co-regulation of genes belonging to the same biological process or molecular pathway. One of the most flexible and powerful gene set tests is the ROAST method in the limma package. ROAST uses residual space rotation as a sort of continuous version of sample permutation. Like permutation tests, it protects against false positives caused by correlations between genes in the set. Unlike permutation tests, it can be used with complex experimental design and with small numbers of replicates. It is the only gene set test method that is able to analyse complex “gene expression signatures” that incorporate information about both up and down regulated genes simultaneously.

ROAST works well for individual expression signatures, but has limitations when applied to large collections of gene sets, such as the Broad Institute’s Molecular Signature Database with over 8000 gene sets. In particular, the p-value resolution is limited by the number of rotations that are done for each set. This makes it impossible to obtain very small p-values and hence to distinguish the top ranking pathways from a large collection. As with permutation tests, the p-values for each set may vary from run to run.

This talk presents Fry, a very fast approximation to the complete ROAST method. Fry approximates the limiting p-value that would be obtained from performing a very large number of rotations with ROAST. Fry preserves most of the advantages of ROAST, but also provides high resolution exact p-values very quickly. In particular, it is able to distinguish the most significant sets in large collections and to yield statistically significant results after adjustment for multiple testing. This makes it an ideal tool for large-scale pathway analysis.

95 2015 AGTA Conference

THANK YOU TO The organising committee for the 2015 AGTA Conference would like to take the opportunity to thank the following organisations for their OUR GENEROUS valuable contribution, and ask that all delegates please consider these organisations for future SPONSORS business relationships.

GOLD SPONSOR SILVER SPONSORS

Welcome Reception Sponsor Name Badge Sponsor BD is a leading medical technology company that partners with customers and stakeholders Illumina is advancing human health by to address many of the world’s most pressing unlocking the power of the genome; from and evolving health needs. Our innovative genome-wide discovery through to targeted solutions are focused on improving medication validation and beyond. management and patient safety; supporting infection prevention practices; equipping sur- At Illumina, our goal is to apply innovative gical and interventional procedures; improv- technologies and revolutionary assays to the ing drug delivery; aiding anesthesiology and analysis of genetic variation and function, respiratory care; advancing cellular research making studies possible that were not even and applications; enhancing the diagnosis of imaginable just a few years ago. infectious diseases and cancers; and sup- These revolutionary tools for DNA, RNA and porting the management of diabetes. We are protein analysis are enabling rapid advances more than 45,000 associates in 50 countries in disease research, drug development and who strive to fulfill our purpose of “Helping all the development of molecular tests in the people live healthy lives” by advancing the clinic. quality, accessibility, safety and affordability of healthcare around the world. In 2015, BD www.illumina.com welcomed CareFusion and its products into the BD family of solutions.

www.bd.com

96 2015 AGTA Conference

Barista Cart Sponsor

As pioneers in sequencing with a rich Keynote Speaker Sponsor heritage in diagnostics, Roche is focussed on developing novel sequencing technologies QIAGEN is the leading global provider of and products to support innovation in clinical Sample to Insight solutions to transform research applications. biological materials into valuable molecular insights. QIAGEN sample technologies isolate We are developing assays, and refining and process DNA, RNA and proteins from techniques, to advance sequencing blood, tissue and other materials. Assay technologies that will deliver truly technologies make these biomolecules differentiated clinical value. visible and ready for analysis. Bioinformatics software and knowledge bases interpret Our current products include NimbleGen’s data to report relevant, actionable insights. wide range of exomes, gene panels and Automation solutions tie these together in custom designs for target enrichment in DNA, seamless and cost-effective molecular testing RNA and methylation applications. We are workflows. Changing Science and Changing Lives. www.qiagen.com www.lifescience.roche.com

Keynote Speaker Sponsor

DNAnexus accelerates the development and delivery of genomic medicine with a global network for sharing and managing genomic data and tools. DNAnexus provides a technology infrastructure and toolset that is optimized for addressing the challenges of security, scalability, and collaboration, for organizations that are pursuing genomic- based approaches to health, in the clinic and in the research lab. www.dnanexus.com

97 2015 AGTA Conference

THANK YOU TO OUR TRADE EXHIBITORS

Trade Booth 5 Trade Booth 8

Agilent is a global leader in target enrichment Agena Bioscience is a San Diego, CA based solutions for next-generation sequencing and life sciences and clinical diagnostics company genomic microarrays. By synthesizing custom that recently acquired the Bioscience business of Sequenom, Inc. and is now offering the complex mixtures of long oligonucleotides, MassARRAY® System. The system is a highly SureSelect and HaloPlex target enrichment sensitive, quantitative method for nucleic acid solutions enable researchers to identify detection via mass spectrometry for high- genomic regions of interest for focused, cost- throughput genotyping and mutation profiling effective variant profiling, with workflows that for cancer and other disease research, generate sequencing-ready libraries in just companion diagnostics, pharmacogenomics, one day. For more information visit Agilent molecular blood group typing, epigenetics, Genomics. clinical genetics, agrigenomics, and molecular sample identification for bio-banking. www.genomics.agilent.com www.agenabio.com Trade Booth 9

Trade Booth 7

Thermo Fisher Scientific is the world leader in serving science. Our mission is to enable our customers to make the world healthier, cleaner and safer. From Partek provides software for next generation Applied BiosystemsTM gold standard sequencing, microarray, and qPCR data anal- Sanger sequencing to simple and scalable ysis and visualization that empowers biolo- Ion Torrent™ next-generation sequencing, gists to find biological meaning within their we have a platform to meet your needs. data. With an intuitive interface designed for Achieve fast, accurate, and affordable researchers, Partek’s complete solution takes results with optimized applications, high- raw data to statistical and pathway analysis quality consumables, and simplified analysis that allows true multi omics integration. solutions. www.partek.com www.lifetechnologies.com

98 2015 AGTA Conference

Trade Booth 10 Trade Booth 12

DNA Genotek provides high-quality biological sample collection, stabilisation and preparation products for human genetics, GeneWorks is an independent Australian microbiology and infectious disease company supplying leading technologies applications. The company’s products protect to Australian and NZ researchers. Unique and stabilise multiple sample types for in Australia, we supply both products and long-term storage at ambient temperature genomics services from our Adelaide facility. to ensure the highest quality results for Products include KAPA kits and Sage genetic analysis and testing. Due to their Science Pippin instruments for NGS library reliability and ease-of-use, DNA Genotek’s preparation, Agena MassARRAY for SNP products are used by thousands of academic, genotyping and cancer profiling and LGC biotechnology, diagnostic, agriculture, and Genomics KASP genotyping technology for other leading institutions around the globe. agrigenomics. Our services include custom SNP genotyping, primer and probe synthesis, www.dnagenotek.com siRNA and gene synthesis. www.geneworks.com.au Trade Booth 13

Trade Booth 11

OnQ Software brings you the best-of-breed laboratory software solutions with the next Macrogen is the current market leader in bio- generation of web-based LIMS, streamlining technology services in Korea, providing pivotal workflows, automating complex and mundane research tools to a global client base. The tasks, freeing up staff to tackle more import- company provides high quality Next Gener- ant work in genomics research and pathology. ation Sequencing and Sanger’s sequencing, Clarity LIMS from GenoLogics was built microarray, oligo synthesis and bioinformatics specifically for genomics life sciences re- services at competitive prices. Using only the search and mass spec laboratories, providing most up to date technology available on the easy-to-use sample tracking and instrument market, Macrogen is able to facilitate large integration, preconfigured workflows and scale evolutionary studies, whilst still taking reporting for QC and regulatory compliance, care to provide smaller scale investigations, matching the needs of your lab staff, collabo- customised to the client’s needs. rators and clients. www.macrogen.com www.onqsoft.com.au

99 2015 AGTA Conference

Trade Booth 14 Trade Booth 16

Gene Target Solutions will be showcasing: Australian Genome Research Facility Ltd (1) The latest generation of NGS Library (AGRF) is Australia’s largest provider of Preparation Technology allowing the end- genomics services and solutions. AGRF user to develop high complexity libraries has laboratories in Brisbane, Melbourne, (exceptional coverage of AT- and GC- rich Adelaide, Sydney and Perth, each providing a regions) from limiting amount of material gateway to a national network of state-of-the- (PCR-free from 10 ng gDNA) and (2) Award art facilities, technology and expertise. From winning Technology currently used by Illumina microbes to plants, animals and humans, to QC NGS preparations. AGRF provides a full range of services across the entire biological spectrum. We provide www.genetargetsolutions.com services to academia and industry with clients from bioscience, environmental science, biomedicine and agricultural biotechnology.

Trade Booth 15 www.agrf.org.au

Trade Booth 17 & 18

Illumina is advancing human health by un- locking the power of the genome; from ge- nome-wide discovery through to targeted validation and beyond. As pioneers in sequencing with a rich her- At Illumina, our goal is to apply innovative itage in diagnostics, Roche is focussed on technologies and revolutionary assays to the developing novel sequencing technologies analysis of genetic variation and function, and products to support innovation in clinical making studies possible that were not even research applications. imaginable just a few years ago. We are developing assays, and refining These revolutionary tools for DNA, RNA and techniques, to advance sequencing technolo- protein analysis are enabling rapid advances gies that will deliver truly differentiated clinical in disease research, drug development and value. the development of molecular tests in the Our current products include NimbleGen’s clinic. wide range of exomes, gene panels and www.illumina.com custom designs for target enrichment in DNA, RNA and methylation applications. We are Changing Science and Changing Lives.

www.lifescience.roche.com

100 2015 AGTA Conference

Trade Booth 19 Trade Booth 21

Scientifix Australian-owned since 1998.

Clontech excels again with the introduction Millennium Science is a specialist distribution of The SMART-Seq v4 Ultra Low Input RNA company established in 1999 specifically to Kit for Sequencing offering the most sensitive source, supply and support high technology mRNA-seq kit for single cells and ultra-low solutions to the Australasian Life Science inputs. Clontech’s SMARTer Single Cell market. The company is well established RNA-Seq kits are now recognised as the as a leading distributor of instrumentation Gold Standard. Also on offer are the SMART and reagents for the research market with ChIP-Seq libraries from DNA inputs > 100 pg recognition of performance and business and Guide-it CRISPR/Cas9 Genomic Editing practice excellence by the global companies products. represented, including Fluidigm, Pacific Biosciences, Affymetrix and more. Macherey Nagel DNA/RNA purification and size selection for NGS libraries. www.mscience.com.au Takara Bio leading Japanese manufacturer of quality PCR, qPCR and RT-PCR enzymes. Trade Booth 22 Laboratory equipment. www.scientifix.com.au

Trade Booth 20 The Ramaciotti Centre for Genomics, located at at UNSW, is a not-for-profit provider of genomic services. We enable access to state-of-the-art, cutting edge technologies at affordable prices and deliver data of the highest quality. Our technology suite includes long- and short-read next-generation Genesearch is home of the e•Freezer and sequencing, single-cell analysis, microarrays sole Australian distributor of New England and Sanger sequencing. Our professional Biolabs and Cell Signaling Technology team of scientists has many years of products. Visit our booth for information experience and provides researchers with a on the NEBNext library preparation kits personalized service from design through to for WGS, RNA-Seq, small RNA and ChIP- downstream analysis. Seq, as well as rRNA depletion, mRNA enrichment, library quant, DNA repair , and www.ramaciotti.unsw.edu.au epigenetics kits. www.genesearch.com.au

101 2015 AGTA Conference

Trade Booth 23 Trade Booth 25-26

BD is a leading medical technology company that partners with customers and The Garvan Institute’s Kinghorn Centre for stakeholders to address many of the world’s Clinical Genomics is a centre of excellence most pressing and evolving health needs. Our for clinical genomic research and diagnostics. innovative solutions are focused on improving We aim to advance the use of genomic medication management and patient safety; information in patient care by delivering and supporting infection prevention practices; interpreting genome sequences for research equipping surgical and interventional and clinical purposes. We provide cost- procedures; improving drug delivery; effective access to genomic technologies and aiding anesthesiology and respiratory care; data interpretation, and our experts in medical advancing cellular research and applications; genetics, pathology and bioinformatics enhancing the diagnosis of infectious proactively engage with stakeholders to diseases and cancers; and supporting the realise the benefits of genomic medicine in management of diabetes. We are more Australia. than 45,000 associates in 50 countries who www.garvan.org.au/kccg strive to fulfill our purpose of “Helping all people live healthy lives” by advancing the quality, accessibility, safety and affordability Trade Booth 24 of healthcare around the world. In 2015, BD welcomed CareFusion and its products into the BD family of solutions.

www.bd.com

Bio-Strategy is a locally-owned, ISO 9001:2008 accredited distributor, supplying leading-edge technology solutions throughout Australia and New Zealand, from quality renowned suppliers. Our experienced staff are highly qualified and trained.

With a culture of innovation and service, under-pinned by our core values of EXCELLENCE – INTEGRITY – TEAM, we focus on providing the best solutions for our customers, backed by technical support from our Applications Scientists and Service Engineers. www.bio-strategy.com

102 AGTA 2015 DELEGATE LIST 2015 AGTA Conference

State/ First Name Last Name Organisation Country Saud Alarifi King Saud University SAU Saad Alkahtani King Saud University SAU Rose Andrews University of New England NSW Stuart Archer Monash Bioinformatics Platform VIC Nicola Armstrong Murdoch University WA Jonathan Arthur Children’s Medical Research Institute NSW Bernard Atmadibrata Children’s Cancer Institute Australia NSW Kelly Avery-Kiejda Hunter Medical Research Institute and the NSW University of Newcastle Warren Bach Millennium Science VIC Melanie Bahlo Walter + Eliza Hall Institute of Medical Research VIC Laura Baker Garvan Institute of Medical Research NSW Colin Baron Qiagen USA Nenad Bartonicek Garvan Institute of Medical Research NSW Brant Bassam Bio-Strategy VIC Denis Bauer CSIRO NSW Traude Beilharz Monash University VIC Natalie Beveridge University of Newcastle NSW Katherine Bolton University of Newcastle NSW Justin Borevitz Australian National University ACT Anthony Borneman The Australian Wine Research Institute SA Nikola Bowden University of Newcastle NSW Rachel Boyle Auckland University of Technology NZL Timothy Budden University of Newcastle NSW Magdalena Budzinska Centenary Institute NSW Murray Cairns University of Newcastle NSW Vicky Cameron University of Otago NZL Dan Catchpoole Children’’s Hospital at Westmead NSW Ross Chapman Hudson Institute of Medical Research VIC Fadi Charchar Federation University VIC Hsiufen Chua Partek SGP Aaron Chuah Australian National University ACT Megan Clarey Illumina VIC Nicole Cloonan QIMR Berghofer Medical Research Institute QLD Joe Copty Kinghorn Centre for Clinical Genomics NSW Susan Corley Systems Biology Initiative UNSW NSW Christopher Cowled CSIRO VIC Mark Cowley Garvan Institute of Medical Research NSW

103 AGTA 2015 DELEGATE LIST 2015 AGTA Conference

Patrick Danoy Roche Diagnostics NSW Cecilia Deng The New Zealand Institute For Plant & Food NZL Research Limited Marcel Dinger Garvan Institute of Medical Research NSW Evan Dodds Illumina VIC Marina Donskoi Mater Health QLD Evan Doods Illumina VIC Alexander Drew ANZAC Research Institute NSW Todd Druley Washington University School of Medicine USA Janette Edson Queensland Brain Institute QLD Richard Edwards University of New South Wales NSW Rob Elshire The Elshire Group Limited NZL Annabelle Enriquez Victor Chang Cardiac Research Institute NSW Sue Forrest AGRF VIC Alistair Forrest Harry Perkins Institute of Medical Research WA Doug Fowler University of Washington USA Saskia Freytag Walter + Eliza Hall Institute of Medical Research VIC Jenny Fung QIMR Berghofer Medical Research Institute QLD Jonathan Gannoulis OnQ Software VIC Brooke Gardiner University of Queensland QLD Velimir Gayevskiy Garvan Institute of Medical Research NSW Goknur Giner Walter + Eliza Hall Institute of Medical Research VIC Brian Gloss Garvan Institute of Medical Research NSW Dominique Gorse QFAB Bioinformatics QLD John Graham ANZAN NSW Sean Grimmond University of Glasgow GBR Paul Harrison Monash Bioinformatics Platform VIC Karl Herron Agena Bioscience QLD Martha Hinton GeneWorks SA Kate Howell University of Western Australia WA Gyorgy Hutvagner University of Technology Sydney NSW Robyn Jamieson Children’s Medical Research Institute NSW Margaret Jordan Comparative Genomics Centre QLD Ben Kaehler Australian National University ACT Naga Kasinadhuni Australian Genome Research Facility QLD Janet Kelso Max Planck Institute For Evolutionary GER Anthropology Matloob Khushi Children’s Medical Research Institute NSW Steve Kujawa Pacific Biosciences USA Carsten Kulheim Australian National University ACT

104 AGTA 2015 DELEGATE LIST 2015 AGTA Conference

Andrew Kyriazis Custom Science VIC Hyun Jae Lee University of Queensland QLD Stephen Lelivre Thermo Fisher Scientific VIC Aaron Lewis AGRF VIC Ruby CY Lin Asbestos Diseases Research Institute NSW Ryan Lister University of Western Australia WA Alexandra Livernois University of Canberra ACT Maria Lubka-Pathak Kinghorn Centre For Clinical Genomics NSW Jesper Maag Garvan Institute Of Medical Research NSW Jovana Maksimovic Murdoch Children’s Research Institute VIC Vikki Marshall Melbourne Neuroscience Institute VIC J Clark Mason BD Biosciences USA Damara McAndrew Genesearch QLD Mark McCabe Garvan Institute of Medical Research NSW Annette McGrath CSIRO ACT Ken McGrath The Australian Genome Research Facility QLD Prachi Mehta University of New South Wales NSW Timothy Mercer Garvan Institute of Medical Research NSW Andre Minoche Garvan Institute of Medical Research NSW Gisela Mir Peter MacCallum Cancer Centre VIC Dessislava Mladenova Garvan Institute of Medical Research NSW Cath Moore Qiagen VIC Caitriona Murray Ramaciotti Centre for Genomics NSW Serena Nik Zainal Wellcome Trust Sanger Institute GBR Mitchell O’Connell University of California, Berkeley USA Alicia Oshlack Murdoch Children’s Research Institute VIC Chris Ozga OnQ Software VIC Swetansu Pattnaik Garvan Institute of Medical Research NSW John Pearson QIMR Berghofer Medical Research Institute QLD Sarah Peaty Roche Diagnostics NSW Steve Pederson University of Adelaide SA Åsa Pérez-Bercoff University of New South Wales NSW Andrew Perkins Mater Research Institute QLD Greg Peters Childrens Hospital at Westmead NSW Mark Pinese Kinghorn Cancer Centre NSW Nooriyah Poonawala- University of Auckland NZL Lohani David Powell Monash University VIC Joseph Powell University of Queensland QLD Mirana Ramialison Monash University VIC

105 AGTA 2015 DELEGATE LIST 2015 AGTA Conference

Michael Rhodes NanoString Technologies WA Natalie Rickers DNA Genotek CAN Christain Rinke University of Queensland QLD Alan Rubin Walter + Eliza Hall Institute Of Medical Research VIC Neil Saunders CSIRO NSW Torsten Seemann University of Melbourne VIC Jaynish Shah Kolling Institute NSW Haojing Shao University of Queensland QLD Jafar Sheikh Jabbari Australian Genome Research Facility VIC Cheryl Shorten Millennium Science VIC Kirby Siemering AGRF VIC Beth Signal Garvan Institute of Medical Research NSW Cas Simons University of Queensland QLD Helen Speirs Ramaciotti CentreFor Genomics NSW Aaron Statham Kinghorn Centre for Clinical Genomics NSW Peter Sternes Australian Wine Research Institute SA Clare Stirzaker Garvan Institute of Medical Research NSW Andrew Stone Kinghorn Centre for Clinical Genomics NSW Alex Stuckey University of Auckland NZL Yuting Sun Children’s Cancer Institute NSW Alex Swarbrick Garvan Institute of Medical Research NSW Andrew Szentirmay Gene Target Solutions NSW Phillippa Taberlay Garvan Institute of Medical Research NSW David Thomas Garvan Institute of Medical Research NSW Daniel Thomson Garvan Institute of Medical Research NSW Natalie Thorne Melbourne Genomics Health Alliance VIC Amali Thrimawithana The New Zealand Institute For Plant & Food NZL Research Limited Sonia Timmis Bio-Strategy VIC Yash Tiwari Thermo Fisher Scientific VIC Cole Trapnell University of Washington USA Nikki Tsoudis Scientifix VIC Rust Turakulov AGRF VIC Ellen Van Dam Bioplatforms Australia NSW Mark Van der Hoek South Australian Health & Medical Research SA Institute Joris Veltman Genome Diagnostics Nijmegen NLD Nic Waddell QIMR Berghofer Medical Research Institute QLD Claire Wade University of Sydney NSW Matthew Wakefield Life Sciences Computational Centre VIC

106 AGTA 2015 DELEGATE LIST 2015 AGTA Conference

Cindy Waters Integrated Sciences NSW Detlef Weigel Max Planck Institute For Developmental Biology GER Trystan Whang Macrogen Oceania NSW Cali Willet University of Sydney NSW Liam Williams University of Auckland NZL Kokulapalan Wimalanathan Iowa State University USA Tanja Woyke Joint Genome Institute DEU Ruidong Xiang CSIRO QLD Jean Yang University of Sydney NSW Amanda Yip Becton Dickinson Holdings SGP Monika Zavodna University of Otago NZL Ryan Zeppel University of Adelaide SA Mahdi Zeraati Garvan Institute of Medical Research NSW Zong-Hong Zhang Queensland Brain Institute QLD Chenxi Zhou University of Queensland QLD

107 Exhibition Floor Plan MALE TOILETS TOILETS SERVICE ENTRY SERVICE FEMALE TOILETS AGTA PLENARY DISABLES TOILETS 8

AGTA REGISTRATION 5

BOOTH EXHIBITOR 5 Agena Bioscience 7 Partek 8 Agilent Technologies 9 Thermo Fisher Scientific 10 GeneWorks 11 Macrogen Oceania

12 DNA Genotek GALLERIA 13 OnQ Software 14 Gene Target Solutions 15 Illumina Australia 16 Australian Genome Research Facility 17 Roche Diagnostics 18 Roche Barista Cart 19 Scientifix 20 Genesearch 21 Millennium Science 22 Ramaciotti Centre for Genomics MAIN ENTRY 23 Kinghorn Centre for Clinical Genomics 24 Bio-Strategy 25 & 26 BD Biosciences

AGTA 2015 SPONSORS AND EXHIBITORS

On softwarQe