World Magazine Bio IT

FORWARD THIS ISSUE TO A COLLEAGUE

SUBSCRIBE TO BIO•IT WORLD MAGAZINE

Welcome to your complimentary digital edition of BiolIT World magazine

We encourage you to evaluate this issue, share it with your team, and apply for your free professional subscription to request regular delivery. Subscribe. We are genuinely excited about the prospects for this field and the magazine in the coming year. ¯ ¯ We've cemented our position as the flagship publication of Cambridge Healthtech Institute (CHI), and will continue to play a prominent role in hosting and reporting on the best CHI conference throughout the year. We pride ourselves on publishing critical insights and analysis of innovations across the drug discovery pipeline from molecular modeling of popular drug targets to biomarkers that discriminate ¯¯ cancer responders, and from data handling for next-generation sequencing to new strategies for increasing the speed and efficiency of clinical trials. We will work hard to surpass those stories in the coming year, continuing our pursuit of the most critical tools and strategies that epitomize the world of “predictive biology.” We look forward to welcoming you as a subscriber, and hope to engage you with our editorial content, and within our network.

NAVIGATION KEY

Previous/Next Page: mouse click (or keyboard) on left/right arrows to turn page

Contents: takes you to the Table of Contents page

To Zoom: In the browser version, simply click on the page to zoom in, and again to zoom out; in the enhanced PDF, use the zoom in/zoom out buttons in the top or bottom tool bar.

Return to Front Cover: In the browser version, just click the double arrow button in the Nav bar; in the PDF, click on the Front Cover in the Tool bar.

Search: Click on the search button to perform a full keyword or phrase search. You can search a single issue, or your complete archive.

Symyx Notebook The Freedom to Experiment.

Whether it’s on the trail or in the lab, you want the freedom to take new approaches, routes, and paths to your goals. That’s why there’s Symyx Notebook. It’s the only electronic laboratory notebook that can be deployed across the enterprise in multiple scientific disciplines. With Symyx Notebook, research teams share a single application to document, work, collaborate, and speed the experimentation workflow.

Symyx Notebook streamlines the capture of all experimental information and intellectual ideas. Everyday tasks such as data capture and note taking are optimized and automated. All of which gives you the time and freedom you need to experiment—and get back to doing science.

To learn more, visit www.symyx.com/notebook6

Cambridge Healthtech Media Group www.bio-itworld.com

Indispensable Technologies Driving Discovery, Development, and Clinical Trials JAN. | FEB. 2009 • VOL. 8, NO. 1

HAVING A BLAST: IMAGING INFORMATICS WITH HHMI’s GENE MYERS 14 THE RETURN OF HIGH- THROUGHPUT SCREENING 20 PREDICTING eCLINICAL TRENDS IN 2009 28

Programming at Pfizer’s BBC David Cox, the new controller of target generation for Pfizer 17

Get Better Results With Oracle

oracle.com/goto/healthsciences or call 1.800.ORACLE.1

Life at Lilly real people doing extraordinary things

Lead the next generation of drug discovery informatics

Eli Lilly and Company is a leading, innovation-driven pharmaceutical corporation with a firm commitment to help people live longer, healthier and more active lives by making breakthroughs in medicines and treatments. To find the next generation of drugs, we know we need to use next generation technologies and employ the best of the biopharmaceutical industry. The Integrative Computational Sciences (ICS) group at Lilly Singapore Centre for Drug Discovery (LSCDD), provides state-of-the-art computational solutions to enable the global efforts of drug discovery, translational medicine and tailored therapeutics at the post genomic era. We study the genetics base of complex diseases and develop novel algorithms, data analysis methods, simulation models and tools for drug discovery, biomarkers identification, epigenetics research, population stratification and prediction of dose response. ICS integrative analyzes solutions, and biological interpretations of complex multi-dimensional data are applied in various therapeutic areas such as Oncology, Diabetes, Neuroscience and Cardiovascular diseases. As an ICS team member with strong analytical and scientific insight, you will have a global impact on the future of personalized medicine: ‘The Right Drug, at The Right Dose for The Right Patient at The Right Time’!

Lilly Singapore is expanding and the ICS group is looking for candidates in the following positions: • Bioinformatics Manager • Sr. Cheminformatics Scientist • Sr. Statistical Geneticist • Principal Statistician • Sr. Bioinformatics Scientist • Scientific Liaison The successful candidates will work closely with their bioinformatics, statistics and software engineering peers at ICS and will collaborate with biologists, chemists, geneticists and physicians at Eli Lilly. We work closely with Lilly System Biology and Drug Discovery Research teams at Singapore as well as with the Discovery Informatics organization and the Global Discovery Statistics group in Europe and USA. Minimum requirements: Preferred Experience: • Ph.D. in Statistical Genetics/Bioinformatics/ • Industry experience in Biotech/Phamaceutical/Drug Computational Biology/Biostatistics/Biophysics/ Discovery Cheminformatics or a related discipline • Experience in the analysis of large data sets such as • 3 years’ post-graduation experience microarrays, sequencing, proteomics and imaging • Background in Oncology / Diabetes / Cardiovascular data Neuroscience • Strong publications record • Statistics: SAS/R/S-Plus/Partek • Demonstrated learning agility • OS: Unix/Linux, Windows • Excellent communication and multidisciplinary • Databases: Oracle/SQL/mySQL collaborative skills • Scripting: Perl, Shell For more information and online application, please visit www.lscdd.lilly.com.sg/lscdd/careers.

Contents [01–02•09]

14 21 28

Feature Story Economic Development 32 BioQatar: A Splash in High Throughput Screening the Desert The tiny Arabian principality is lur- Researchers are finding value, productivity, ing scientific entrepreneurs, venture capitalists, and American universities and hits in the much maligned technology. 21 to spur biotech development.

In Every Issue Up Front Computational 5 Illumina Wins This 8 IT-Enabled Personalized Development Round on Points Medicine 17 Target Programming at First Base Illumina sequencers are Personalized medicine proponents Pfizer’s BBC behind the 3 genomes published discuss what wisdom health-IT David Cox leads one of Pfizer’s recently in Nature, but new solutions brings to the movement. “biotech-like” research units. are on the horizon. BY KEVIN DAVIES 10 2009’s Future for Pharma 42 Super Models at IBM Summit James Greenwood talks about pipe- Russell Transcript The IBM Modeling lines, stem cells, and a free market IT/Workflow & Simulation Summit showcases the health care system. 25 SC08 Showcases HPC strengths and weaknesses of modeling. BY JOHN RUSSELL 12 Coming Event: The 2009 Muscle Systems Bio•IT World Expo IBM retains top spot on Top500 list, 6 Company Index two systems top the petaflop, and The event features breakthroughs in 6 Advertiser Index quad-cores take over. semantic technologies, high perfor- 34 New Products mance computing, and more. 40 Educational 13 Conducting Clinical Trials in Clinical Research Opportunities Emerging Markets 28 2009 Predictions and Insights Outlook A new report high- Resolutions lights the market for clinical trials in Waife and Sirabella see “on-shoring” Eastern Europe. BY AL DOIG as a key trend for the New Year. SPECIAL ADVERTISING SECTION 12 Briefs 29 Streamlining Regulatory eCTD Publishing Computational Biology Automated eCTD submissions can Lab Automation, 14 Imaging Informatics save companies time, money. Eugene Myers prepares to 30 Three Key Ingredients LIMS, and Data BLAST off again. to Cost-Effective 16 Building Integrative Biology Clinical Systems Integration at Boehringer Ingelheim CDISC ODM, Web-services, Begins on page 36 David de Graaf finds a new home and the right business model and focus in integrative biology. smooths clinical trials.

[4] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com Cover photo by Seth Affoumado

First Base

ment regarding how far sequencing technology has come. “No one in this room would have thought that in a single publica- Illumina Wins tion in Nature, there’d be effectively four genomes published.” (Kahn counted both the tumor and normal genomes in the Wash U. study.) This Round Illumina also announced the sale of 11 more GA instruments to the Wellcome Trust Sanger Institute, bringing their tally up to 36. It’s not that the GA instruments have dramati- on Points cally (or even incrementally) greater throughput or accuracy than SOLiD, but the Sanger Institute already has the workflow, informatics pipeline, and trained personnel in place for that platform. KEVIN DAVIES Of course, AB—now Life Technologies following the merger with Invitrogen —isn’t taking this lying down. Kevin McKernan n August 2005, on the eve of the publication of the first (see “The Drive for the $1000 Genome,” Bio•IT World, May commercial next-generation sequencing technology, 454 2007) anticipates the SOLiD version 3.0 will reach a human Life Sciences founder and Chairman Jonathan Rothberg genome cost of around $10,000 early this year, though how I preemptively declared: “We won the race. Everyone may quickly those prices can be translated into the field remains to not be happy with that, but we are.” be seen. An AB salesperson at ASHG said that the three major Three years on, the publication of not one but three com- advantages of its 3.0 SOLiD system were flexibility, accuracy, plete human genomes in the same issue of Nature last No- and throughput. AB recently announced it was placing six vember effectively marks the end of the latest chapter in the SOLiD instruments with the Broad, and eight more to the BGI. sequencing saga. These included the first African genome, led by David Bentley and colleagues at Illumina UK; the first Asian What Happens Next? genome, produced by Jun Wang and colleagues at the Beijing Now, all eyes are on what’s next for next-gen sequencing. Genomics Institute (BGI) in Shenzen, China; and the first can- British hopes have shifted from Cambridge to Oxford, where cer genome, that of a deceased leukemia patient, produced by Oxford Nanopore Technologies is making impressive strides Elaine Mardis, Rick Wilson, and colleagues at the Washington on both the chemistry and engineering fronts, details of which University School of Medicine Ge- should be forthcoming later this year. nome Center in St Louis. Six thousand miles away, Complete Genomics will soon The common thread in those present details of its first assembled human genome and start publications was the platform— taking orders for its $5,000 genome service. (CEO Clifford Illumina’s Genome Analyzer (GA), Reid will keynote the 2009 Bio•IT World Expo—see p. 11). which bought British company Pacific Biosciences recently published the first examples of its Solexa in 2006 for a cool $600 real-time single-molecule sequencing technology. Helicos re- million. Steady advances in read cently laid off one third of its workforce, but hopes for stronger length, throughput, and accuracy sales under new CEO Ronald Lowy, who takes over from Steve Scott Kahn gave Illumina the edge over its Lombardi. With many other companies lying in wait, 2009 rivals. “We won,” says John Milton, the former Solexa head of should see plenty more fascinating twists and turns. chemistry now with Oxford Nanopore Technologies, even while acknowledging that Applied Biosystems’ (AB) SOLiD instrument might be of higher quality. Clive Brown, Solexa’s former Best Practices Last Call head of bioinformatics also now at Oxford Nanopore, agrees. On April 28, Bio•IT World will host a dinner to an- “We won. Enough of the boxes were ticked in terms of efficiency nounce and celebrate the winners of the 2009 Best and performance and it’s won. It’s clearly well ahead of every- Practices competition during our annual Expo in one else.” Howard Hughes Medical Institute investigator Tim Boston (see p. 11). The entry window remains open Harris, an authority in single-molecule biology who left Helicos for a few more weeks, so we encourage any group— last April, confirms, “The Illumina guys have clearly won this academic, industry, user, vendor—to submit examples round, there’s no doubt.” of partnership and innovation enhancing drug develop- Illumina marked the genome milestones with a press con- ment, biomedical and clinical research. Full details and ference at the American Society of Human Genetics (ASHG) guidelines, along with the final entry deadline, can be convention in Philadelphia last November. The company’s CIO, found at www.bio-itworld.com/bestpractices. Scott Kahn, said the simultaneous publications made a state-

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 5 ]

Company Index ®

Indispensable Technologies Driving Discovery, Development, and Clinical Trials

23andMe ...... 8 FDA ...... 29 NIH ...... 32 454 Life Sciences ...... 5 Genedata ...... 22, 23 Novartis ...... 12, 20, 42 EDITOR-IN-CHIEF Abbot Labs ...... 12 Genentech ...... 17 Oak Ridge National Laboratory . . 25 Kevin Davies (781) 972-1341 Accelrys ...... 23, 34 Georgetown University ...... 32 Oppenheimer & Co...... 7 [email protected] Affymetrix ...... 17 GlaxoSmithKline ...... 21 Oracle ...... 8 Allen Brain Institute ...... 11 Google ...... 25 Otto-von-Guericke-University MANAGING EDITOR Amgen ...... 12 Harvard Medical School ...... 11 Magdeburg ...... 15 Allison Proffitt (781) 972-1345 Applied Biosystems ...... 5 Helicos ...... 5 Oxford Nanopore Technologies . . . . 5 [email protected] Argonne National Laboratory . . . . 25 Hewlett-Packard ...... 8 Pacific Biosciences ...... 5 Ariadne ...... 12 HHS ...... 8 Panasas ...... 25 ART DIRECTOR Beijing Genomics Institute ...... 5 Howard Hughes PerkinElmer ...... 21, 22 Mark Gabrenya (781) 972-1349 Bioinformatics Organization . . . . . 12 Medical Institute ...... 5, 14 Perlegen ...... 17 [email protected] BioTeam ...... 11 IBM ...... 25, 42 Pfizer...... 11, 16, 17, 20, 28, VP BUSINESS DEVELOPMENT Biotechnology Industry Illumina ...... 5, 8, 12 ProtoLife ...... 22 Angela Parsons (781) 972-5467 Organization ...... 10 ImClone ...... 12 RainDance Technologies ...... 34 [email protected] BlueArc ...... 25 Imperial College, London ...... 33 REvolution Computing ...... 12 Boehringer Ingelheim ...... 16 Improvision ...... 22 Sanger Institute ...... 5 VP SALES – WESTERN US, Broad Institute ...... 12 Institute for Computational Siemens Medical Solutions ...... 8 CANADA, EUROPE, PACIFIC RIM Carnegie Mellon ...... 32 Medicine at Johns Hopkins . . . 42 Sony ...... 25 Alan El Faye (213) 300-3886 Celera ...... 14 Invitrogen ...... 5 Southern Research Institute . . . . 20 [email protected] Clinical Data Interchange Isilon Systems ...... 25 Symyx ...... 34 REGIONAL SALES MANAGER – NEW ENGLAND, Standards Consortium ...... 30 Johnson & Johnson ...... 11 Texas A&M ...... 32 NORTH EASTERN US, SOUTH EASTERN US, MIDWEST, INDIA Complete Genomics ...... 5, 11 Knome ...... 8 University of California Berkeley . . 14 Kay O. Christopher (860) 693-2991 Convey Computer Corp ...... 25 LabCorp ...... 8 University of California [email protected] Cornell University ...... 32 Lawson Security ...... 34 San Francisco ...... 18 Cray ...... 25 Life Technologies ...... 5 University of California SENIOR DIRECTOR OF MARKETING deltaDOT ...... 33 Locus Pharmaceuticals ...... 12 San Diego ...... 25 & OPERATIONS, PUBLICATIONS Department of Energy ...... 25 Los Alamos National Laboratory . 25 University of Cincinnati ...... 23 Joan A. Chambers (781) 972-5446 DiscoveRx ...... 22 McKesson ...... 8 University of Rochester ...... 42 [email protected] Eli Lilly ...... 8, 11 Merck ...... 12, 23 Vertex ...... 42 PROJECT/MARKETING MANAGER EMEA ...... 7 Merlin Ventures ...... 32 Waife & Associates ...... 28 Lynn Cloonan (781) 972-1352 Entelos ...... 42 Merrimack Pharmaceuticals . . . . 42 Washington University School of [email protected] Excalibur ...... 32 Microsoft ...... 12, 25 Medicine Genome Center . . . . . 5 ExecuPharm ...... 28 Millennium ...... 20 Wellcome Trust Sanger Institute . . . 5 ADVERTISING OPERATIONS COORDINATOR Fabra University...... 42 Navigenics ...... 8 Wyeth ...... 12 Matt Gross (781) 972-5481 [email protected]

PRODUCTION MANAGER Tom Norton (781) 972-5440 Advertiser Index [email protected]

Advertiser Page # Advertiser Page # Contributing Editors Barnett Educational Services ...... 19 Lilly Singapore Centre for Drug Discovery ...... 3 Michael Goldman, Karen Hopkin, www.barnettinternational.com www.lscdd.lilly.com.sg ...... Deborah Janssen, John Russell, Bio-IT World 2009 Best Practices Award ...... Cover 3 Molecular Medicine Tri-Conference 2009 ...... 9 Salvatore Salamone, Deborah Borfitz www.bio-itworld.com/bestpractices www.Tri-Conference.com Ann Neuer, Tracy Smith Schmidt Bio-IT World Conference and Expo 2009 ...... 26-27 Oracle ...... Cover 2 www.Bio-ITWorldExpo.com www.oracle.com/goto/healthsciences Advisory Board Consumer Connectivity 2009...... 33 Post-Approval Summit ...... 15 www.worldcongress.com/connectivity www.postapproval.org Jeffrey Augen, Mark Boguski, GenoLogics ...... 38-39 Symyx ...... 36-37, Cover 4 Steve Dickman, Kenneth Getz, www.genologics www.symyx.com/notebook6 Jim Golden, Andrew Hopkins, LabVantage ...... Digital 7 Caroline Kovac, Mark Murcko, www.labvantage.com John Reynders, Bernard P. Wess Jr. This index is provided as an additional service. The publisher does not assume any liability for errors or omissions. Cambridge Healthtech Institute PRESIDENT VOLUME 8, NO. 1 Phillips Kuhl Editorial, Advertising, and Business Offices: 250 First Avenue, Suite 300, Needham, MA 02494; (781) 972-5400 Bio•IT World (ISSN 1538-5728) is published bi-monthly by Cambridge Bio Collaborative, 250 First Avenue, Suite 300, Needham, MA 02494. Bio •IT World is free to qualified life science professionals. Periodicals postage paid at Boston, MA, and at additional post offices. The one-year subscription rate is $199 in the U.S., $240 in Canada, and $320 in all other countries (payable in U.S. funds on a U.S. bank only). POSTMASTER: Send change of address to Bio-IT World, P.O. Box 3414, Northbrook, IL 60065. Canadian Publications Agreement Number 41318023. CANADIAN POSTMASTER: Please return undeliverables to Station A, P.O. Box 12, Windsor, ON N9A 6J5. Contact Information Subscriptions: Address inquires to Bio-IT World, P.O. Box 3414, Northbrook, IL 60065 (888) 835-7302 or e-mail [email protected]. [email protected] Reprints: Copyright © 2009 by Bio-IT World All rights reserved. Reproduction of material printed in Bio •IT World is forbidden without written 250 First Avenue, Suite 300 permission. For reprints and/or copyright permission, please contact the YGS group, 1808 Colonial Village Lane, Lancaster, PA; 717-399-1900 Needham, MA 02494 X 125 or via email to [email protected].

[6] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

International Business Machines Corporation in the United States, other countries, or both. Linux is a registered trademark of LabVantage Solutions, Inc. © 2007. All Rights Reserved. Sapphire™ is a trademark of LabVantage Solutions, Inc. IBM, the IBM B DISCOVERY · DEVELOPMENT · BIOBANKING · FORMULATION · QUALITY MANAGEMENT

A WORLD OF LABORATORY INFORMATION MANAGEMENT OUT OF THE BOX Linus Torvalds in the United States, other countries, or both. usiness Partner emblem and System x are trademarks or registered of

Sapphire™ Suite: The Best Combination of Functionality, Flexibility and Technology Successful organizations can no longer depend on one-dimensional Sample systems. As your business grows, so must your systems and their functionality. Mgm't

Sapphire Suite Pl LabVantage redefi nes the “S” in LIMS with the -- a zero-footprint ate 't & A gm rray Mg Spec. M and confi gurable-off-the-shelf Laboratory Information Management Suite. m't Clinica Mgm't l Data M & Prot. Sapphire takes you beyond the world of traditional LIMS by consolidating gm't Stability additional functionality typically found in multiple laboratory systems into Project & Study Mgm't Adv. Storage & Logistics y Custody gurabilit & Gen one product platform across the global enterprise. Sapphire is the best st Confi ealogy Tr Va acking on A zati dv. combination of functionality, fl exibility, & technology in one solution. nali Re tio por ltina ting Mu Confi gure your success with the Sapphire™ Suite at: Secure Sys. eData Inst. & ration www.labvantage.com/suite Entry Integ Keyword: OutOfBoxLIMS

AUSTRIA • BELGIUM • CANADA • CHINA • DENMARK • ESTONIA • FINLAND • FRANCE • GERMANY HONG KONG • INDIA • JORDAN • LATVIA • LITHUANIA • LUXEMBOURG • MALAYSIA • NETHERLANDS NORWAY • OMAN • QATAR • SAUDI ARABIA • SINGAPORE • SOUTH AFRICA • SWEDEN • SWITZERLAND TAIWAN • THAILAND • UNITED ARAB EMIRATES • UNITED KINGDOM • UNITED STATES

Up Front News

IT-Enabled Personalized Medicine Personalized medicine proponents discuss what smarter, cleverer, and simpler. Frederick Lee, product manager for wisdom health-IT brings to the movement. personalized medicine and genomics for McKesson agreed. IT has to take a long view of the development, he said. The in- BY ALLISON PROFFITT personalized medicine must be shown in frastructure has to be in place before per- OSTON—Personalized medicine the real world and enthusiasts must show sonalized medicine can really take off. Lee can learn a lot from health-IT, that it saves more money than it costs. compared personalized medicine to the said David Brailer at the annual Health-IT had to prove long term benefits early automotive industry. After all, there BHarvard Partners Personalized in concrete amounts, and personalized couldn’t be an explosion in car sales until Medicine conference.* Brailer, the former medicine must do the same. there were roads and fueling stations. HHS “health czar,” proposed that health- Finally, Brailer told the audience that IT is about five years ahead of personal- they had to show doctors how personal- Managing the Data ized medicine politically and culturally, ized medicine would work for them in Linda Avey, co-founder 23andMe, liked and that the head start offers wisdom for their clinics and hospitals. Just like train- the automotive analogy. Representing personalized medicine. ing physicians to use an EMR, doctors will consumer genomics offerings, Avey and need to be helped through the change, colleagues gave their “wish lists” for IT and shown how to implement personal- support for personalized genomics and ized medicine in their daily practice. personalized medicine. Avey said that Illumina is the road and Vendor Perspective LabCorp the fueling station for 23andMe. Rising to Brailer’s challenge, a roundtable Right now we are a Prius, she joked, but of IT vendors discussed the solutions hope to become a semi full of data head- required to enable personalized medi- ing to the hospital. Avey encouraged pro- cine. Translating data into knowledge is viders to make their electronic medical daunting, and vendors must craft a toolkit records compatible with personal health to serve personalized medicine’s needs, records such as the offerings from Google said Eiry Roberts, VP transitional phase and Microsoft to allow patients to transfer development, Eli Lilly. genomic data to their providers. Neil de Crescenzo, senior VP and Dietrich Stephan, CSO Navigenics, general manager, health care and life imagined a “magical algorithm” would sciences, Oracle , said his priority is to take into account all of the genetics as David Brailer is applying health-IT lessons learned to personalized medicine. advance open IT standards and speed the well as the temporal exposures of an in- ability to search images. In the company’s dividual and print out a report, but until Brailer said personalized medicine new 12G platform, it has presented DNA then, Stephan believes we must still do proponents need to recognize that the as a data type, and he hopes that will reasonable risk assessments with known full value will not initially be understood speed personalized medicine research. risk factors and focus on better commu- by everyone, certainly not the public. Janet Dillione, CEO, health services, nication of those assessments and better The health-IT movement had to educate Siemens Medical Solutions, pointed clinical-decision support for doctors. experts to communicate with the public. out that IT vendors don’t own the data. David King, CEO of LabCorp wel- “We viewed health-IT as a campaign,” Medical knowledge has to be in the public comes the fueling station comparison and Brailer said. Just as his staff campaigned domain. The semantic web will help link notes that there’s an important difference to Congressmen, governors, and anyone patient, genomic, and medical data, and between well-known and well-validated that would listen, so he believes “per- service oriented architecture will be es- results. Validation of results, analyses, sonalized medicine needs that level of a sential to managing the data volume and and even the sources that interpret the war waged. It’s not going to drift into the advancing personalized medicine. data is essential to successfully applying American consciousness.” Lionel Binns, business development genomics data to health. Brailer also laid the burden to prove executive, health and life science, Hewl- Jorge Conde, CEO of Knome noted superiority on the innovator. Early health- ett-Packard, said that the main lesson for there isn’t a central database of genom- IT efforts asked to be proven wrong, and health-IT is to stay simple. IT can’t lead ics data. The cost of sequencing is falling that doesn’t work, he said. The benefits of the campaign into personalized medicine, quickly, but understanding the data isn’t *Harvard Partners Personalized Medicine: A Value Prop- it can only provide tools, and they should progressing as rapidly, and tools to under- osition; Harvard Medical School, November 12-14, 2008. be not bigger, faster, and better, but stand the data should be a priority. x

[8] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

Up Front News [IN CONVERSATION ]

debate, articulate their fervent desire to cure cancer, diabetes, Parkinson’s and Al- 2009’s Future for Pharma zheimer’s, then turn around the next day and vote in ways that make it very difficult James Greenwood talks about pipelines, stem cells, for us to attempt doing so. We’re going to be very focused on trying to demonstrate and a free market health care system. to the new administration and Congress that the process might begin with an NIH grant, but it then requires a patent Former Congressman James Greenwood system that works, policies creating tax has been the Biotechnology Industry incentives for investing in startups, and Organization’s (BIO) president and CEO a well-resourced, well-staffed FDA with since January 2005, and continues to be a good, confirmed Commissioner at the widely viewed as a leader on health care front. It also requires a set of realistic ex- issues in many Washington, D.C. circles. pectations with regard to drug risks and Late last year, Alissa Poh had the oppor- benefits, instead of Congress constantly tunity to speak with Greenwood about his beating up the FDA because of a safety thoughts on the new administration, and signal. It will require follow-on biolog- its impact on pharma/industry. ics legislation that provides a sufficient period of data exclusivity so companies Bio•IT World: Are you hoping to convince will make investments in new products; the new leadership that this is no time as well as reasonable reimbursement to to cut drug prices? the manufacturers so they get a return on Greenwood: If those in Congress who their investments. are most determined to reduce drug prices are wildly successful, managing Tell us more about your hopes for BIO to achieve European-level pricing, that and the industry, under Obama. might save $50 billion annually—about UPI/BRIAN KERSEY On my more optimistic days, I’m hopeful James Greenwood the R&D budget for private sector drug that the kind of change President Obama discovery. I fear that focusing on driving whacking away at drug pricing. It’s insuf- talked about during his campaign isn’t down prices will significantly impede the ficient, and can ultimately be counterpro- just rhetoric, but that it really causes the ability of companies to use their product ductive in terms of research dollars. There political system in Washington to start revenues for development, and dissuade are four things we’d like in health care rethinking the way it does things. For investors from not only Big Pharma, but reform: universal access, quality, afford- instance, change doesn’t mean that once small startup biotechs. The best way to ability, and innovation. Too frequently, the inauguration is over, both sides draw cut health care costs is to reduce the in- however, that last piece is forgotten. The the same swords and have the same fights, cidence of chronic disease, which could Europeans solved the access and afford- with a slightly different outcome. Maybe save us $150 billion or more. And we can ability problems, but there’s been a price change means that before we start engag- do this by helping biotechs develop new to pay in quality and innovation. Our big ing in these battles, we have a conversation products that prevent such illnesses from challenge is whether we in the U.S. can about how to achieve win-win outcomes, expressing themselves to begin with. create a better system. how to work more collaboratively and with less conflict. As examples, Medicare Then how would you address the ques- What are your thoughts on reversing the Part D price negotiations, and re-impor- tion of dried-up pipelines? embryonic stem cell research policy? tation—are these simply calendar dates, Big Pharma has had some difficulties We’ll certainly applaud reversal of that like Groundhog Day, where we have to with their pipelines, yes. We think the policy, but we’ll also use the occasion wage war with the same heated rhetoric solution lies in healthy, mutually sustain- as a teaching moment. It’s extremely and accusations? Couldn’t we discuss the ing relationships with biotechs and small important that members of Congress real safety issues and whether importation companies, who are particularly available understand that simply allowing the NIH actually accomplishes anything in terms of right now for collaborations of various to fund embryonic stem cell research in affordability and access? kinds—M&A, investments, licensing universities does not produce therapeu- On my more cynical days, I think we’ll deals—given the credit crunch. tic/diagnostic products for patients. That probably just repeat the same routines. translation is quite far down the road But I think we owe it to ourselves as a na- So BIO supports preservation of the free from academia; it’s done by the private tion to take President Obama at his word market health care system? sector and our companies. What has that he wants to do things differently—so It’s short-sighted to think the place you always been frustrating is to have these we need to be available and encouraging begin with in health care reform is by members, in the context of the stem cell about a different approach. x

[10] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

Coming Event: The 2009 Bio•IT World Expo The event features breakthroughs in semantic technologies, high performance computing, network modeling, genomic medicine, and more.

The 2009 Bio•IT World Conference & Consumer Genomics Expo takes place April 27-29 in Boston. Information Full details of the program are available eThree years ago, Mark Bo- online (www.bio-itworldexpo.com) but guski keynoted the Expo in his below are some of the editors’ highlights: capacity as director of the Allen Brain Institute. After a stint at What’s Hot in IT Novartis, Boguski is now build- One of the most engaging, informative ing a new company, Resound- and downright popular presentations ing Health, in the personal at Bio•IT World Expo is traditionally genomics space. Boguski will Chris Dagdigian’s no-holds-barred re- discuss the challenges facing view of the key trends and pet peeves he consumers and service provid- experiences day in day out as a BioTeam ers in the new world of personal consultant working in academic and in- genomics. dustrial data centers. Dagdigian is one of $5,000 (see, “Complete Genomics Tar- eHealth Solutions | Track 6 | April 28 the regular contributors to Bio•IT World’s gets ‘The First $1000 Genome,’” Bio•IT ‘Inside the Box’ column. Last year, Dag- World, November 2008). Rather than digian’s talk was the highlight of the IT selling instruments, Complete Genomics infrastructure track. This year, Dagdigian is offering a next-generation sequenc- delivers the opening keynote. f ing service model, with big pharma very Keynote Lecture | April 27 much targeted as a key customer demo- graphic. In one of his first presentations to a scientific audience, Reid will discuss Semantic Technologies technology advances and his vision for the The semantic web may not be a reality future of genome sequencing in medicine just yet, but Pfizer’s Ted Slater gave a and pharmaceutical research. fascinating presentation at last year’s Keynote Lecture | April 29 Bridging Pharma & IT conference on his group’s promising implementation of semantic technologies. His talk in Track Improving Performance 2—“You’re using your computer wrong: A full-day workshop on high-performance Semantics in pharma R&D”—is highly computing trends and applications recommended. will explore grid and cloud technolo- Track 2 | Informatics for Genomic gies, and a range of HPC applications Medicine | April 28 including next-gen sequencing. Speakers include the excellent John Halamka (CIO Har- Coming Soon: vard Medical School), The $5000 Genome Reece Hart (Ge- e Last October, nentech), and Cliff Reid (CEO, Giles Day Complete Genom- (Pfizer BBC), ics) made the as- as well as se- tounding claim nior IT staff that his company from Johnson would sequence & Johnson 1,000 human ge- and Eli Lilly. nomes in 2009 Workshops | for a net cost of April 27

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 11 ]

Up Front News Briefs

CLUSTERED R The 2009 Bio•IT World Expo REvolution Computing is working with Microsoft to bring the “R” tion suite for sharing clinical and trans- statistical language to Microsoft’s lational medicine data. Speakers include new Windows HPC Server 2008. director Ken Buetow, who accepted the REvolution’s RPro is an optimized, Editors’ Choice award at last year’s Best validated, and commercially sup- Practices event (see, “Connecting the ported distribution of R (see, Bio•IT Cancer Community caBIG Time,” Bio•IT World, July 2008, p. 30). RPro pro- World, August 2008) grams can now utilize the multiple Tracks 1, 5, 6 | April 28-29 microprocessors inside Microsoft computer clusters, automatically scaling to leverage whatever Making Sense of microprocessor count is available. Complex Data The RPro module that enables clus- eEric Schadt enjoyed a remarkable year ter computing, ParallelR, runs in in 2008. The Merck Research Laboratory interactive mode on Windows HPC scientist led a sizeable multidisciplinary Server 2008. REvolution has also team to some truly groundbreaking ar- integrated RPro with the Windows ticles in top journals integrating various HPC Server 2008’s scheduler, allow- forms of genomic data, including associa- ing users to queue RPro batch jobs. tion studies and expression data to build models for target and drug discovery (see, ILLUMINA INTERPRETATION Benjamin Franklin “Eric Schadt’s Integrative Approach to Ariadne has completed a partner 2009 Award Predictive Biology,” Bio•IT World, Octo- program with Illumina to provide Who will join the ranks of distinguished ber 2008). advanced integration and inter- winners of the Benjamin Franklin Award, Keynote Lecture | April 28 pretation capabilities for data including Sean Eddy, Michael Ashburner, generated from Illumina’s iScan Ewan Birney, Lincoln Stein, James Kent, and BeadArray Reader platforms. and Michael Eisen? The winner of the Working IT The coupling of Ariadne’s Pathway 2009 award will be introduced by the A selection of five workshops on the first Studio to Illumina’s next gen- Bioinformatics Organization president, day of Bio•IT World Expo offer a little eration software, GenomeStudio Jeff Bizarro. something for everyone. Topics range Data Analysis Software, provides Franklin Award | April 28 from the application of Wikis by life researchers with a state-of-the-art science organizations and drug safety toolset to quickly analyze expres- informatics to visualization in medical sion data for physiological changes, Computational Chemistry informatics and advances in molecular cell signaling and other significant An outstanding line-up of speakers dynamics and ligand docking. biological changes. features in Track 4 on computational Workshops | April 27 chemistry, including the Broad Institute’s SPECIFIC LABELING David DeCaprio, Locus Pharmaceuticals’ Amgen and ImClone met with Jeff Wiseman, and Journal of Chemical FDA in late December to ask the Information and Modeling editor Anton agency to relabel their cancer drugs Hopfinger, as well as informatics leaders Vectibix and Erbitux in response to from Wyeth, Novartis, and Abbot Labs. clinical trial data showing that the Track 4 | April 28-29 drugs work in patients with a normal KRAS gene, but not in patients with a mutated version. Labeling caBIG Time with drugs with this informa- The work of the Cancer Bioinformaticsf tion would be tantamount to FDA grid will be prominently displayed approval of the drug for a certain throughout the conference program. group of patients. About 40% of Presentations include the deployment colon cancer patients are believed of a grid infrastructure across 540 to have a mutated KRAS gene (see, medical centers and a review of Bio•IT World, April 2008, p. 8). caBIG’s clinical trials applica-

[12] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

Insights | Outlook

clinical trials recruiting patients at the end of 2007. A second group of European countries have either decided Clinical Trials against EU membership or have applied for and are awaiting EU community acceptance. These countries, either because of historical affiliations or applicant status requirements, have also integrated EMEA standards into their policies and regu- in New Europe lations governing clinical trials. Among the countries in this category are Bosnia, Croatia, Serbia, Switzerland, and Turkey. Bosnia, Croatia, and Turkey were hosting 194 of all global clinical trials recruiting patients at the end of 2007. The third group of countries consists of members of the CIS and includes Russia, Ukraine, and Belarus, who together were hosting 393 of all global clinical trials recruiting patients at the AL DOIG end of 2007. These countries have been updating their policies and regulations governing clinical trials in order to attract clini- s of early 2008, there were 50,629 clinical trials cal trial projects. Related to these efforts, an analysis within a ongoing globally ( Oppenheimer & Co.), up by 1.3% report of FDA inspection outcomes of clinical trials conducted from 2007. While the United States still hosts the in these countries indicates that their performance is equivalent largest number of clinical trials, there are more than to—and in one case better than—their “Western” counterparts. 150 other site countries around the globe. Growth in numbers of clinical trials conducted has led to Opportunities in New Europe Aincreased competition among sponsor companies and CROs The countries covered in this report differ widely with regard to find clinical sites and to recruit clinical investigators and to population size, distribution and quality of health care, treatment-naïve patient populations. forms of health care delivery, and availability of state-of-the- Europe offers a vast new opportunity for clinical trials as a art facilities and trained investigators. However, in an result of the collapse of the Soviet Union. Dozens of sovereign increasingly competitive environment for patients and countries have emerged, and gone is the political investigators, many of the countries of dividing line between East and West Europe. Now the New Europe offer attractive options. these countries join the countries of Western Europe Among these attractions are large, as well as the Commonwealth of Independent States In an increasingly treatment-naïve populations eager to (CIS) to offer a spectrum of clinical trial options. competitive participate in trials to receive advanced A new publication from Insight Pharma Reports, treatment. In certain cases, the European Clinical Trial Site Options: An Insider’s environment, population density of these treatment- Analysis, provides a penetrating review and summa- many of the naïve patients is within acceptable travel tion of opportunities for clinical trial programs in 44 distance to a country’s most prestigious countries, with contributions from 25 locally based, countries of New university and medical school complex, clinical trial specialists. where medical staff and administrators With the dramatic geopolitical shifts that have Europe offer are eager to conduct clinical trials, and occurred over the past two decades, Europe now attractive options. where costs are low. extends from Iceland in the West to the Ural Moun- What becomes clear from the report tains in the East. The expansion of NATO and the desire for EU is how heterogeneous this landscape is on a number of dimen- membership have been major drivers. This report organizes sions. For example, even with the EMEA policies as a common the countries within this enlarged geographical area into three denominator, most countries have layered their own sovereign groups: EU members, non-EU members, and CIS members. view to aspects of the clinical trial process. This is best illus- The growth in EU member states has carried with it the trated by the variations in clinical trial approval procedures and integration of EMEA regulations governing clinical trials, required documents described in the report for each country. expanding clinical trial options beyond Western Europe. Coun- These differences are likely to decrease as efforts at harmoniza- tries such as Bulgaria, Cyprus, Czech Republic, Estonia, Hun- tion continue to evolve. However, what is encouraging from the gary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slove- report’s analysis is the degree to which the EMEA policies are nia are EU Member States and were hosting 2.5% of all global the scaffold for clinical trial regulation across Europe.

Further Reading: European Clinical Trial Site Options: An Insider’s Analysis, by Pavle Vukojevic, MD. Published by Insight Pharma Reports, November 2008, Al Doig is general manager of Insight Pharma Reports. He can www.InsightPharmaReports.com. be reached at [email protected].

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 13 ]

Computational Biology

Imaging Informatics Eugene Myers prepares to BLAST off again.

BY JOHN RUSSELL you a readout for how much there he gratuitous use of BLAST here is in some gamish of cells. Here will probably make Eugene (Gene) you can go into individual cells and Myers cringe—not that the How- see where it is, and see how much T ard Hughes Medical Institute of it there is, and what the distribu- (HHMI) investigator doesn’t want to re- tion is. This is much higher dimen- peat his past success, this time in imaging sional data, and qualitatively much informatics. Myers and colleagues at the more interesting information. Of NIH developed BLAST in the 1990s to course it’s harder to get, but it is cope with the rising tide of sequence data. getting progressively easier to do After writing an audacious proposal on this.” the theoretical shotgun sequencing of the human genome, he lobbied to join Craig Data Handling Venter at Celera, where as VP informatics The key problem, as Myers puts it, Eugene Myers research he put his computational chops is that, “People have the ability to on the line. generate lots of data but they don’t have protein interactions and behavior and Having returned east from UC Berke- the software infrastructure for handling systems biology. “It’s going to create op- ley in 2006 to become one of the founding it. Basically people are rolling their own portunities that don’t otherwise exist and investigators at HHMI’s Janelia Farm and in a lot of cases they’re asking for create knowledge that it would be very, (see, “Marshall’s IT Plan for Janelia help.” The situation is not unlike genom- very hard to get by looking at an array of Farm,” Bio•IT World, October 2006), ics in the ’80s, he says—a rapidly emerg- numbers that tell you kind of generally Myers is again wading into a data flood. ing technology with software develop- how [gene] expression went up or down But instead of dealing with the puny 3 ment playing catch up. in a collection of thousands of cells.” billion base pairs of human DNA, now he “That’s kind of why I’ve entered the is tackling the 4.2 trillion voxels required arena,” says Myers. “There are new com- IT Challenges to image a mouse’s brain. Affordable and putational challenges and problems, some Myers knows something about wrestling improved imaging technology is promis- of which can be solved in part by using with data challenges from his Celera days, ing to light up the molecular landscape of existing methods in the imaging litera- so he’s not fazed by what lies ahead. “We’re living systems with far-reaching impact ture, but they are not solved adequately going to generate data sets that are larger on basic research, drug development, and by those techniques. Moreover nobody is than the Celera data set,” he says. The in the clinic. in a position to deploy these things in the data will fall into three broad categories: Myers points to two major technol- contexts that are arising.” 1) Looking inside cells to observe proteins ogy jumps in microscopy in recent years. “I kind of placed my bet. I’ve sat myself and various elements; 2) looking at collec- First is the ability to capture information down right next to the biologist… that’s tions of cells to see how they are organiz- digitally. “CCD (charge-coupled device) really the only way that it works. If you try ing and interacting; and 3) interpreting detectors that are cheap and high resolu- to do it remotely, it doesn’t work very well. video or other imaging data that indicates tion and sensitive are a recent develop- You have to really immerse yourself in the the behavior of the resulting systems like ment. We’ve only had really great ones for pipelines that are producing the data.” a mouse moving or a fly flying. the last five or so years,” he says. Second, Myers hopes to hit another home Imaging an entire mouse brain rep- Myers points to the development of fluo- run—like BLAST—but the work is chal- resents 4.2 trillion voxels, or terabytes of rophores that can be expressed. “We have lenging. “I’m having more trouble finding raw data requiring interpretation. “That’s the entire genomes of many organisms so the killer app, but I’m convinced that a big number,” jokes Myers. Interpreting that we can make any part of the genome a couple of killer apps will ultimately those data begins with a large Beowulf glow and any part of protein that gets pro- emerge,” he says. Less certain is whether cluster. “The unusual thing about our duced by a genome, we can make glow.” this field will become as big as genomics, cluster, unlike say a Google-type cluster, is “Now, we have the opportunity to ob- because it is somewhat more technology- that our machines are very high memory. serve in vivo, in situ, in the cell directly, intensive and cross-disciplinary. We have more expensive machines and the expression of genes. This is much dif- Myers believes imaging will be a big- the reason for that is that images are very ferent than [microarrays], which just give ger driver of knowledge in cell biology, large and you want to operate on large

[14] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

3-dimensional arrays. It’s really kind of a Beyond his own research, Myers sees High-throughput microscopy will en- convenience; we’re buying our way out of a huge opportunities for imaging in diag- able pharma teams to screen the uptake hole rather than really struggling with it.” nostics, while stressing that he’s merely a of a chemical into thousands of cells, with Myers applied a similar buy-their-way- scientist, not a physician. He cites work of a digital readout as to whether the cell is out strategy a decade ago at Celera. “We Walter Schubert, Otto-von-Guericke-Uni- dividing. were buying a 64-gigabyte memory, which versity Magdeburg (Germany), Molecular Myers’ major ongoing research proj- was actually one of the largest commercial Pattern Recognition Research Group, ects, both involving C. elegans, are devel- memory you could buy. Now we routinely Institute of Medical Neurobiology, from oping a single-cell expression atlas of C. have quite a few processors with those big a few years ago, who was staining cells elegans, and studying the biophysics of memories on them. The smallest memory for molecular targets. “As soon as you put mitosis. Further interests include develop- on any of our machines is 8 gigs.” Moving down this marker for basically the pres- ing a complete light level atlas of the fly’s the data is another headache. “You’ve got ence of an antibody, you could see that the brain with its complete developmental to get 4.2 terabytes out of the disk system antibody in one case had penetrated into trajectory. His group is also working to the various processors. So you have to the skin layer and in the other it hadn’t.” on behavioral scans—recording mouse move huge volumes and so distributed file In cancer, Myers says it will be possible whiskers while the animal is being stud- systems are very important to use.” to mark a cancer with certain reagents ied electro-physiologically, and trying to Myers plans to release tools using and look for markers. “It will be about build a high-throughput microscope to open source. “It’s been extremely tricky the distribution of that protein and its capture high-dimensional, entire vol- to go commercial anyway in the scientific presence in certain cells that you can’t get umes of brains to understand stochasti- enterprise,” he says. “It’s really hard to find from an expression array. It’s that high- cally the fine-grained flow of neuronal an edge where you have exclusivity and dimensional aspect of actually seeing the information. customers willing to pay the requisite distribution and the pattern in an actual Asked to divulge more about the mi- overhead. I’m more interested in getting histological context that will tell us what croscope, Myers demurs. “I want to keep stuff out there.” the disease will be.” my edge for a while longer.” x

Learn strategies and best practices for demonstrating safeetyt , quq aalityy, effeectctivveness, and value through patient rer gig sstries, phasa e IV studies, risk managageme ent prograams and quality inniti iatives. Join industry leadderss, impllementet rs andd otheh r sts akeholdeersr in the prp eeminentn cono fef rencce on thee evolviing area off Posst-t Approvall proogrg amms.s

For moree infn oro maatit on, viv sit us at www.postapproval.org oro contaactt us ata 617.621.6426.

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 15 ]

Computational Biology

But we’re not proud—we’ll take expression data, protein array data, whatever Building Integrative Biology we can get. What we really want to do is integrate all of those data sets to help us at Boehringer Ingelheim understand the biological context of our disease and our targets.” An early project David de Graaf finds a new focus in integrative biology. is in early safety and toxicology, identifying assays that will enable de Graaf to “start to look at de-risking our portfolio very early on for safety issues.” BY JOHN RUSSELL disease—helping to identify targets and Using natural language processing Boehringer Ingelheim (BI), the Ger- the modalities to be deployed to affect and other technologies, de Graaf’s team man family-owned biopharmaceutical them. He will marshal a package of bio- is already delivering relevant project in- conglomerate, is making a (modest) bet logical tools, from animal models to text formation to scientists. As for the systems on integrative biology at its Ridgefield, mining and mathematical models that biology effort, de Graaf is starting with a Conn., facility. Leading the new effort put the targets in context. His group will small team before working with academic as VP of biotherapeutics and integrative also support both NBEs (new biothera- and biotech partners to expand their biology is David de Graaf, who until last peutic entities) and NCEs (new chemical capabilities, aiming for about ten people summer was Pfizer’s director of systems entities) as the start of a lead optimization over the next year or so. He projects a biology at the Research process. similar size for the e-biology team. Technology Center (RTC). He sees excellent De Graaf sees integrative biology as That division is now synergy between “sub- “the ability to take data sources from a under the control of Pfiz- systems biology”—the number of different places and integrate er’s new Biotherapeutic understanding of dis- them to help us understand biological sys- and Bioinnovation Cen- crete pieces of signal tems better.” That includes understand- ter (BBC), led by Corey transduction—and an- ing the target in context, as well as the Goodman, based in San tibody therapeutics. downstream consequences of manipulat- Francisco (see page 17). “You can very quickly ing a particular target. “Part of what we Ironically, de Graaf will be know a lot more about integrate is not only data but also models, reporting to Phil Vickers, antibody structures whether it’s the human disease model in the former RTC chief (see, because the space in primary cells or animal models, we need “Pfizer’s Persuit of Technol- which they move is just to integrate all of those.” ogy,” Bio•IT World, April so much less complex He continues: “Sometimes the only 2006) who now heads BI than intracellular tar- thing we need is a bit more context and a

research in the U.S. Vick- FILE PHOTO/MARK GABRENYA gets,” he says. Not that bit more understanding of what’s affected ers is “here to create a drug David de Graaf it’s going to be easy, by a particular modality or your target hunting culture,” says de Graaf. however; de Graaf admits, “we’re starting and we are able to move ahead with a de- At a time of economic crisis, particu- at zero.” cision. Sometimes we need to understand larly for big pharma, de Graaf relishes the things in much greater detail, and we are opportunity to work for a successful De-Risking the Portfolio going to make models that are predictive family-owned organization and focus on Details of the new integrative effort of biology, of human disease biology.” a long-term mission. Having relocated to are scarce as the effort ramps up, but De Graaf says he’ll know when his ef- southern Connecticut, one of de Graaf’s De Graaf will have a group focused on forts have been successful when the first key challenges is how to attract the same biomarkers, building on BI’s existing two or three projects hit the clinic. Ideally, talent available around the large aca- strengths in genomics, and genetically en- he would like to see external pharma part- demic centers. gineered organisms. Other core strengths ners looking at BI’s portfolio in a couple of De Graaf has been assigned respon- include protein purification and protein years and say, “Wow, your portfolio looks sibility for products as well as for new expression. different; your portfolio looks exciting.” science and enabling technologies. “A big “We’ve just started to build an elec- Today, most pharma portfolios look de- attraction was the ability to go and get tronic biology group—e-biology—which pressingly alike. Says de Graaf: “What we something very concrete out that would is led by Will Loging. He’s got a big focus want to do is appropriately differentiate have the stamp of my organization on it,” on predictive models, predictive biology in areas and start to take risks there in he says. in general.” The e-biology effort is based order to be able to move into new areas De Graaf will be working closely with on “overlaying data sources from a num- and not always be competing.” x the two therapeutic areas—immunology ber of different places, including text Editor’s Note: The full interview with de Graaf is avail- and inflammation, and cardiovascular mining, as a way of getting to those data. able at www.bio-itworld.com.

[16] BIO•ITWORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

Computational Development

Programming at Pfizer’s BBC David Cox leads one of less problematic biotherapeutic discovery tivities by putting human genetics on the and development, and possibly the emer- front end of that process. “Human genet- Pfizer’s “biotech-like” gence a new paradigm that is adopted ics has never been very successful by itself, throughout the Pfizer organization. so it needs human cell-based models and research units. it needs systems biology approaches to be Watching the BBC really useful.” Cox says the TGU will put a The BBC is actually a federation of small, premium on integrating the genetics, cell- BY JOHN RUSSELL independent biotech-like research units based models and systems approaches “to tart with a person who should be that aims to harness the strengths of bio- maximize that synergy.” sick but isn’t. Use DNA variation to tech and pharma to discover and develop After several years at Perlegen, Cox hunt for broken genes that are keep- biotherapeutics through clinical proof- was drawn to the Pfizer opportunity by Sing this person well. Set up a vigor- of-concept. One of those units is Rinat, “the ability to plug human genetics into ous hypothesis generation and killing a spinout from Genentech with cutting- a situation that can impact medicine. It’s machine to mercilessly winnow through edge antibody technology expertise. remarkable what’s happened in human the options. Make drugs informed by the Another is CovX in San Diego, which is genetics over the past five or ten years. rare surviving hypothesis and race them pursuing an emerging technology, uniting But the actual impact on health out- through proof of concept. the therapeutic attractiveness of peptides comes, to date, has been pretty pathetic.” To a considerable ex- Human genetics can play a role in picking tent, this is what Pfizer’s treatment options and use them more year-old Biotherapeutics effectively, Cox says. “But the way our and Bioinnovation Center society is set up, it’s really hard to have (BBC), which recently broke that make commercial and regulatory ground on a headquarters sense. What I realized is that it’s going in the Mission Bay district to be really hard doing it in a diagnostics of San Francisco, is trying way, so why not apply human genetics to do with its Target Genera- make new medicines for unmet medical tion Unit (TGU), led by needs? That is what attracted me to this David Cox. Like his new opportunity.” boss and BBC president Corey Goodman (see, “Can Target Audience Academics Save Pharma?” The TGU approach will start with an Bio•IT World, November important clinical question and outcome, 2007), Cox has a sparkling then use genetics as a way of linking that academic resume. He was outcome to a specific biological pathway. the former co-director of the As an example, Cox offers this intriguing Stanford University Genome Cox left Perlegen for target generation at Pfizer. question: Why do some diabetes patients Center before co-founding the have such a high risk of dying from heart Affymetrix off-shoot Perlegen (see, “Tak- with the beneficial clinical properties of attacks? No current diabetes drugs im- ing Data Storage to Infinity and Beyond,” antibodies. pact the incidence of heart attack, so what Bio•IT World, January 2003). Much of The third unit is comprised of the RTC biological pathway or target is the key to the TGU is located at the BBC Research and Coley in Düsseldorf, Germany, which dealing with that medical need? Technology Center (RTC), in Cambridge, is building Pfizer’s capability in RNA This is where the human genetics Mass. interference. “Each of these research comes in. Cox’s team is starting with If drug finding is still largely searching units are focused on a novel and growing unusual diabetes patients that for some for needles in a haystack, Cox is using a technology area, each unit harnessing reason do not develop cardiovascular genetics magnifying glass to find them. its small size, unique culture, and entre- disease—patients that should have devel- But this is not the whole formula. The preneurial drive to discover and develop oped the side effect but didn’t. “By doing BBC seeks the nimbleness of biotech and breakthrough science in the most produc- DNA sequencing quite globally on a large a new, more effective approach to collabo- tive way,” says Cox. number of genes, we’re looking to see if rating with academia (see, “BBC’s Univer- While those sites will feature cell-based there is evidence of rare DNA variations sity Challenge”). The result, Pfizer desper- and animal in vivo studies, Cox says the that may be the reason why these diabet- ately hopes, will be faster, less expensive, TGU was set up to complement those ac- ics didn’t get cardiovascular disease,” says

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 17 ]

Computational Development

Cox. “We’re looking for the broken genes in the people who didn’t get sick.” In many cases, Cox suspects, the muta- tion will have partially reduced the effect of the gene. “It doesn’t give you a home run, but it parachutes you in specifically to the biology you’d like to look at.” From there, TGU is positioned to study cell-based models to study the effects of knocking out that gene. “Here’s where the systems biology comes in: you perturb the system with that very focused initial gene and now you say, based on the output of that perturbation, are there better things within that aspect of biology that you should be looking at then the initial gene that you used?” Cox says there are two very different

ways of thinking about systems biology. / THE SAN FRANCISCO CHRONICLE BRANT WARD The approach he’s advocating starts with Bay Area biologist Corey Goodman, who leads Pfizer’s BBC, speaks at a groundbreaking cer- one part of the universe and expands out. emony at the company’s new site last August. The other collects the data and reverse engineers how everything works. “They’re cal outcomes that you care about.” Quality Control both absolutely fine intellectual constructs The goal is to use various approaches The success of the BBC will ultimately be and they both undoubtedly contribute re- to draw links between DNA variations judged, Cox says, by “how many things ally useful information, but one is much and clinical outcomes. “Without under- that we put into humans in clinical trials more focused than the other,” he says. standing the mechanism, the chances show positive proof of concept in patients. Cox sees important roles for in silico that things are going to work in humans Period. There’s no intermediate success.” biology, although he says the problem with as a drug are very unlikely.” Cox intends He expects to have some idea in a year or ’omics approaches is that “they aren’t tied to “keep pushing at the mechanism until two, but as for proof of concept in humans, to the fundamental clinical outcomes that we are confident that we have enough “it’s unrealistic to think that we will have we care about. The reason for looking at understanding that when we put that in the answer in less than three. But it better the DNA variation is because you can as- humans, it’s actually going to have the ef- be a lot faster than ten years!” sociate that in a casual way with the clini- fect we want.“ Cox does not see Pfizer’s size and bu- BBC’s University Challenge

Pfizer is focusing on six priority areas based on the per- particular purpose and that we’d like to work with you to ceived likelihood of success and unmet medical needs: get this particular answer.’” That’s quite a departure from oncology, Alzheimer’s disease, schizophrenia, pain, rheu- the traditional way of doing things, says Cox, because in matoid arthritis, and diabetes. The BBC is no different, but general big pharma has been reluctant to divulge its specific will initially focus on diabetes and oncology. interests and strategies. “I think that leveraging what’s out A key component will be a directed emphasis on external in the biomedical world in an effective way is really, really collaborations. Historically, Cox says, “pharma has given important.” money to academics but often hasn’t given any real detailed Cox mentions a new relationship with UCSF as an knowledge about what they are trying to accomplish. And example. Pfizer entered a three year collaboration with the academics are happy to take the money but they aren’t the California Institute for Quantitative Biosciences (QB3), really interested in the questions the pharma industry is headquartered at UCSF. The idea is that UCSF scientists can interested in and they use it to do whatever they want.” propose ideas based on their research. If a Pfizer scientist Cox says Pfizer and the BBC are going to find the scien- feels that idea could be interesting, those two scientists can tists with the right expertise to complement its internal submit a joint proposal to a review team. “The new collabo- strengths, but in a much more directed fashion. “We’re ration intimately links the scientific talent of UCSF with the going to say, ‘We’re really interested in your models for this extraordinary expertise of Pfizer,” says Cox. J.R.

[18] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

The Path to Drug Development, Regulations, reaucracy hindering his group’s goals. “We have a tremendous opportunity to and Industry Trends show how this can work,” he says, arguing that the BBC represents a truly innovative operating model both for Pfizer and the pharmaceutical industry in general. The BBC will work closely with the rest of Pfizer R&D “to maximize platform line, therapeutic area and technology expertise to drive the development of high-value compounds across both organizations.” He foresees plenty of interaction between the two groups. “The BBC will take compounds through proof of concept. We Purchase your copy(ies) today have our own clinical unit. But for all of for your colleagues, research the phase III trials and all of what needs staff members, and professionals to get leveraged from big pharma, that’ll involved in the management and be part of the Pfizer organization. So we conduct of clinical research. leverage the strengths of the big organization but we stay separate enough that we To order, call 1-800-856-2556 or can use a biotech approach for the first visit the ‘Publications’ section on parts of research and development.” www.barnettinternational.com Maintaining a sense of independence for the BBC is important. “We want to preserve the unique cultures of these sites PAREXEL’s Bio/Pharma- and maintain their entrepreneurial spirit,” ceutical R&D Statistical says Cox. “The notion is to have groups Sourcebook 2008/2009 that you can get everybody together in a small auditorium and talk to everybody without a microphone—they’re all fo- The U.S. Drug Approval Trends and cused on the technology, and it becomes Yearbook 2008/2009 really exciting.” There will still be a priority on reaching the go/no-go decision early. “The New Drug Development: motive to kill projects is very high for us A Regulatory Overview 2008 because we expect most of our ideas are going to be not very good,” Cox admits. “No matter how good our ideas are, if we do the right science, most of them are Expediting Drug and Biologics going to go away. That’s just the way sci- Development: A Strategic ence works.” Approach Systems and in silico biology approaches will help scrutinize ideas that appeared promising but based on kinetic Biologics Development: A parameters or mass action, tell the TGU Regulatory Overview not to bother to do a single experiment because it’s dead on arrival. “Lowering clinical attrition is the key,” says Cox. “If we can put biotherapeutics into the clinic Regulatory Reporter Newsletter that have a higher probability of making Barnett International: A division of it to the market to become human drugs, Cambridge Healthtech Institute that will be a win for Pfizer, and a big win www.barnettinternational.com. 250 First Avenue, Suite 300 for human health.” x Needham, MA 02494 USA Phone: (800) 856-2556 Editor’s Note: Read the full interview with David Cox, first published in Predictive Biomedicine, at www.bio- itworld.com www.barnettinternational.com

Feature

High Throughput Screening

Researchers are finding value, productivity, and hits in the much maligned technology. By Vicki Glaser

as high-throughput screening (HTS) suffused drug discovery roviral agent maraviroc (Selzentry), approved by the FDA in 2007, as an example pipelines with “hits” of sufficient quality and quantity to war- of a successful HTS program. In Decem- rant the millions of dollars invested in the technology since the ber 1997, a Pfizer HTS campaign gener- mid-1990’s? That debate has been raging for years. Anemic ated a hit list that included the molecular scaffold destined to become maraviroc. productivity, a surplus of me-too drugs, and costly and embar- Although the drug took nearly ten years to rassing late-stage and post-market failures continue to plague reach market, Bojanic describes that pace as “extremely fast.” Hthe pharmaceutical industry, bolstering critics of HTS who challenge its In fact, HTS has done what it was return on investment. But HTS is evolving; users are learning how to apply designed to do, namely to screen increas- HTS technology for optimal output, and the technologies HTS has spawned ingly large compound libraries and determine which compounds bind to a target. are extending downstream to lead optimization and compound profiling. Most of the problems attributed to HTS lie on either side of the automated screen- The hype that accompanied the in- But given the lengthy drug developing process. As these issues are steadily troduction of HTS in the 1990’s virtually ment cycle, compounds that emerged rectified, HTS grows more robust. Im- assured its failure to meet expectations. from HTS campaigns in the early 2000’s provements upstream of screening have HTS was predicted to revolutionize drug are still traveling through the pipeline. resulted in higher quality compound col- discovery. It did not. In 2004, the percep- Drugs in development today represent lections stored in smaller, more thought- tion of HTS as a failed business strategy “the fruits of our labors 10 to 15 years ago,” fully designed libraries to feed the HTS was celebrated in a front-page story in the says Dejan Bojanic, HTS leader at Novar- engine. With the application of Chris Lip- Wall Street Journal under the headline tis. About 55% of the compounds [now] inski’s Rule of Five and other algorithms, “Drug Industry’s Big Push into Technol- in Novartis’ lead optimization pipeline are “the promiscuous hitters and bad actors ogy Falls Short,” with critics calling the projects that came out of HTS,” he adds. have been removed,” says Lucile White, combination of HTS and combinatorial Bojanic, a biochemical pharmacolo- manager of the High-Throughput Screen- chemistry “an expensive fiasco.” Those gist, has two decades of experience using ing Center at Southern Research Insti- technologies were widely blamed for the HTS for small-molecule drug discovery, tute, an original member of the NIH’s drought in new drug approvals that ac- including previous stints at Pfizer and Molecular Libraries Screening Center companied the first decade of their use. Millennium. He points to Pfizer’s antiret- Network (MLSCN). Additionally, more

[20] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

Retools for the Future

physiologically relevant, functional assays steep learning curve. Combined with diverse, with 10-15 analogs in a series in- are providing a window into the cellular the synergies realized through parallel stead of 50-100 and more chemical series milieu and yielding data more predictive technology advances and strategic shifts represented. Vendors have “dramatically of drug activity in the human body. Cell- in the implementation of screening tools, improved the liquid handling technology based assays are now routinely performed HTS is finally beginning to gain traction. and the detection technology, and we now in a high-throughput screening format But an essential challenge remains—how have much more confidence in the data.” (see, “Hard Cell Made Easier”). to extract, interpret, and apply the hidden “HTS is now seen as a tool rather than Downstream, various labeling tech- nuggets of information from the deluge of a technology,” says Alan Fletcher, VP niques and detection systems are en- raw data that HTS produces? business development at PerkinElmer, abling multiplexing, analysis of drug activity at the single-cell and subcellular level, and characterization of the effects of target binding on signaling pathways TS is a “shotgun and protein-protein interactions, as well H as off-target drug activity. HTS formats approach” most useful with are becoming increasingly prevalent in ‘ secondary assays, enabling earlier studies new targets about which of compound attributes such as ADME/ tox, solubility, or microsomal instability little is known. and prediction of serious adverse drug Lucile White, Southern Research Institute effects such as hERG potassium channel activation. Both HTS itself and access to in vitro ADME/tox and other data earlier Losing its Luster suggesting that its value may be more in in the drug development cycle “are short- “Lots of mistakes were made in the 1990’s,” filtering out inactive compounds rather ening timelines for getting compounds notes Robert Hertzberg, VP screening and than identifying active ones. It soon be- into the clinic,” says White. compound profiling at GlaxoSmithKline came evident that primary screens based Economic realities are also favoring (GSK). Two key problem areas early on on biochemical assays to detect target HTS as cost-cutting efforts and downsiz- were compound collections—insufficient binding offer limited predictability of ing spotlight the advantages of automated diversity and lack of appropriate chemical what would be active in the body. processes, maximizing output while mini- properties—and screening strategies— White describes HTS as a “shotgun mizing manpower needs. The push in sub-optimal assays and problems with approach” that is most useful with new recent years toward assay miniaturization false positives and negatives. targets about which little is known of the has also helped reduce screening costs “By the late 1990’s, we had solved most types of compounds that might bind to and conserve valuable compound stores. of these problems,” says Hertzberg. For them. If the screeners are lucky, they will HTS has reached the backside of a example, chemical libraries are now more uncover several novel structural classes

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 21 ]

Feature

that interact with a target’s active site to up front understanding of the biology to space using data accumulated from previ- pass along to the medicinal chemists. be targeted, the chemical space to be ex- ous screening runs and various statistical One of the first challenges for HTS was plored, and the attributes of the molecule learning algorithms, including neural and miniaturization, to reduce the cost of large to be created. All of this information Bayesian networks. screening campaigns and preserve “the should guide the design of the screening The techniques underlying PDT be- crown jewels”—pharma’s rapidly growing library. Indeed, library quality is more long to the broader field known as Design libraries of synthetic small molecules and important than quantity—a lesson the in- of Experiments (DoE). Traditional ap- natural compounds. This is now routine, dustry learned the hard way. Companies proaches to DoE “relied on dimension and the cost of doing HTS has leveled off. have begun the painstaking task of whit- reduction,” says Packard, identifying and Says Hertzberg: “Most of the industry has tling down their compound collections, exhaustively screening the most relevant settled on a 5 Õl [assay] volume, which purging them of compounds of dubious subspace. PDT eliminates this need. To can be run in a 384- or 1536-well plate.” identity and purity and more thoroughly screen a 1-million compound library, for The cost is about 7¢ per well, and has been characterizing those remaining. example, PDT may apply its modeling for the past couple of years. “In the beginning, when there was techniques to the first 100 compounds— However, “we are approaching the a lot of low-hanging fruit, companies either selected randomly or chosen to limit of how many wells we can cut in a would run a huge number of screens and achieve the best possible coverage of the sheet of plastic,” says White. In the future they did not have to be particularly smart chemical space—and use the results to microtiter plates will give way to microar- about their experimental bandwidth,” predict the next set of 100 compounds, ray type dot arrays on a glass or plastic says Norman Packard, CEO of ProtoLife . and the next set, and so on. surface, she predicts. “Now it is more important to get the big- Then there is the critical issue of The early limitations of HTS led to a gest bang for the buck.” target selection. Bojanic stresses the more realistic view of the technology as ProtoLife’s Predictive Design Technol- ongoing need for more and better target merely a starting point for identifying ogy (PDT) is an intelligent modeling tool validation. As biochemical assays increas- small molecules to progress into medici- that finds optimal hits in complex experi- ingly share the limelight with cell-based nal chemistry programs. “The screening mental spaces through iterative screening, screens, and the initial focus on G-protein part is not what keeps me up at night,” versus a roll-of-the-dice, one-shot screen- coupled receptors (GPCRs) and kinases says Bojanic. A greater concern is: “Have ing approach. At each iteration, PDT’s al- expands to a broader spectrum of target we got the right compounds in our librar- gorithms select the type and complexity of classes, target selection and assay design ies? Have we got the right targets?” models to use for a particular experimen- become an increasing challenge. HTS is not a shortcut, emphasizes tal system based on multiple parameters. “We want to look at protein-protein in- Bojanic, and it cannot replace a thorough The technology models the experimental teractions in a cellular environment, and

Hard Cell Made Easier Sailaja Kuchibhatla, senior VP business development at to improve throughput, cost efficiency, and accessibility. DiscoveRx, describes cell-based assays as “the missing Genedata’s Heyse foresees improvements in capturing the link” between HTS and predictive, functional data. Cell- microscopic image readouts from HCS, which can reveal based assays shed light on the effects of target binding on subcellular activity and physiological responses. In Heyse’s downstream second messengers. With DiscoveRx’s Path- vision, image analysis produces data and information with- Hunter beta-arrestin assays, “we are still looking at GPCRs, out requiring scientists to analyze each image manually. but we are looking in multiple different pathways,” says This information then feeds back into the experimental Kuchibhatla. design process. “We need to automate and make sense of The assays incorporate a target cytoplasmic protein fused all this multidimensional data to understand the activity to a ß-galactosidase (ß-gal) peptide, while another portion and mechanism of action of a compound,” Heyse says. of ß-gal is present only in the cell nucleus. If GPCR activa- “HCS will eventually go plug-and-play,” predicts Perkin- tion causes the target protein to translocate to the nucleus, Elmer’s Richard Eglen, noting growing demand for simpler the two ß-gal peptides can produce an active enzyme and HCS software that can be used by a spectrum of drug dis- generate a signal. DiscoveRx has developed a palette of covery researchers. They want high-throughput assays for assays that encompass more than 150 known receptors and pathway analysis, cell signaling, and evaluating a range 80 orphan receptors. Kuchibhatla describes growing interest of preclinical biomarkers, says Eglen. PerkinElmer has in applying the company’s technology to perform HTS for the strengthened its HCS portfolio with the acquisition of Evotec purpose of compound profiling. and the Opera HCS image analysis platform, and the addi- The growing interest in data-rich high-content, cell- tion of Improvision’s Velocity software, which enables view- based screening has encouraged imaging system vendors ing of live cell images in three dimensions. V.G.

[22] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

we need simple readouts to determine if a compound is affecting a cellular pathway,” says Richard Eglen, PerkinElmer’s HTS Plus a Systems Approach president of bio-discovery. Ion channels are a good example of a new, “rich mine Jeffrey Hermes, Director, Merck Automated Biotechnology (North Wales, Penn.) sees for drug discovery,” notes Eglen, “but they HTS working for Merck in conjunction with a parallel systems biology approach. are difficult to do in HTS.” Bio• IT World: How has the view and implementation of HTS changed in a large Where’s the Beef? pharmaceutical company such as Merck? Despite an improving perception of HTS, Hermes: We continue to see both the promise and the value of HTS every day. The many still question its ability to con- investments we have made in parallel screening technologies are providing us tribute to a knowledge base rather than with an unprecedented breadth and depth of information on candidates across simply produce massive amounts of data. our therapeutic areas of focus. With these technologies we can assay 5,000 to Extracting value from HTS depends on 7,000 candidates in 40-100 different types of biology at the same time. We can data analysis and management strategies also identify novel molecular targets and pathways by “knocking down” and/or and applying those data going forward. “knocking in” gene expression in multiple cell lines using various combinations of Stephan Heyse, head of screening test compounds and perturbing agents. informatics at Genedata, describes HTS as “a big, organized roulette game for Critics argue that HTS has not fulfilled its promise to fill drug discovery pipelines finding actives without the need of a pri- with high value hits. Do you agree? ori knowledge of structures and binding There has been a great convergence of technologies and disciplines in recent motifs.” However, the reliability of early years, which is moving the entire field forward. For example, our high-throughput HTS data was questionable. “Companies biology division works closely with the Molecular Profiling and RNA Therapeutics gathered the hits and then threw away the groups to find new ways to treat common diseases. Through these collaborations, data—they didn’t trust it,” he says. With we are dissecting disease pathways with great precision and identifying novel improved assays come better readouts molecular targets in diseases such as obesity, atherosclerosis, diabetes, and bone and more reliable data, but the increasing disorders. complexity of the data continues to put pressure on the computational toolkit. How do you see the role of HTS continuing to evolve, and will it remain a core HTS data about individual compounds technology in the drug discovery toolbox? and chemical series might prove useful Merck has launched a “poly-target” profiling initiative that uses systems biology not only for hit identification but also at to improve the productivity of drug discovery. Whereas traditional high-through- later stages of lead selection, optimiza- put screening is a linear, empirical process in which multiple drug candidates tion, and preclinical drug development, are screened against one target at a time, Merck has invested in technology that Heyse says. allows us to analyze many critical cellular parameters—including gene and protein Collecting and managing these data expression and metabolites—in parallel. This parallel, systems biology approach throughout the drug development process generates rich information for the simultaneous interrogation of many targets. By makes it easier for team members to share probing multiple cellular functions on an industrial scale and employing sophis- benefits from the information. Heyse ticated data analysis tools, we are transforming high-throughput screening into emphasizes the importance of being high-throughput biology, and we are finding targets we couldn’t find using tradi- able to annotate screening data across tional linear screening. V.G. an organization. In this way, biologists, chemists, and pharmacologists can understand and interpret the data without Research Institute, University of Cincin- quality control,” says Heyse. It contains having to review the history of a screen or nati, to fully integrate Screener into the algorithms that subtract artifacts in the a particular compound series or to look center’s HTS workflow. data. In cell-based assays, for example, up the experimental parameters. As new Screener includes the company’s new cells in the center of a plate tend to grow data emerge, users can add additional Kinetics Analyzer module, which provides differently than those nearer the edges, annotations, create detailed compound tools for analyzing and aggregating time yielding subtle differences in output. profiles, rank compounds, and update series data, in which hundreds of data Screener automatically corrects these their knowledge base. points may be collected from a single well. gradients, and applies a computational, Screener 6.0 is Genedata’s enterprise- The module optimizes parameters and systematic strategy to deal with outliers. wide data management platform for HTS aggregation rules for individual assays “HTS generates as much misinforma- and high-content screening (HCS). Gene- and systematically gathers mechanistic tion as information,” says Frank Brown, data recently signed a collaboration with information from the curve shape. VP and CSO of Accelrys , but iteratively the Drug Discovery Center at the Genome Screener provides “exhaustive data screening more carefully designed com-

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 23 ]

Feature

pound sets aimed at building a structure- with predictive com- activity relationship (SAR) will produce putational modeling 80% higher quality hits and valuable, predic- lies more with the false 1% Active tive information. negatives than with 70% Confirmation rates for HTS campaigns the false positives. The 0.5% Actives are usually below 50%, which Brown ex- typical activity cut-off 60% 0.1% Actives plains as follows. Consider a screen that companies use to de- 3% Actives yields a 1% hit rate (Fig. 1, red line) and a cide which compounds 50% 1% error rate—which Brown says would to move forward to hit be “phenomenal, a best-case scenario.” confirmation is 50%, 40% Only half of the 1% hits will be true posi- mainly because that tives, the other half are false positives. If produces a “man- Active % True 30% the active rate drops to 0.5% (yellow line), ageable workload,” then only 33% will be real hits. A 0.5% Brown says (Fig. 2). 20% hit rate combined with a 2% error rate Shifting the cut-off to generates only 20% correct information 30% would double the 10% and 80% misinformation. workload, and a 20% If one was to use the full dataset gener- breakpoint would qua- 0% ated screening a large random compound druple the workload. 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% library, Brown says you could not build a But Brown argues that Assay Error Rate model capable of predicting which hits new information con- are actually false positives and which tinues to be revealed Fig 1: Confirmation rates for HTS campaigns are usually below 50%, even with a “phenomenal, best-case scenario” error rate “inactives” are actually false negatives. as the activity cut-off of 1%. The range of activities represented is too approaches 20%-30%, diverse—“the data is all over the place,” primarily because of eliminating the false he says. However, if one was to use the negatives. 0 data points generated from a selected Although expanding the search area 50% set of about 1,000 compounds to build to include all the false negatives may be a model for designing subsequent, itera- cost-inefficient, Brown says the use of tive screens, that model would accurately informatics and computational modeling predict about 75% of the false positive (in- systems makes it feasible to broaden the active) compounds and 80% of the false search to the 20% to 30% mark. Scite- negatives (actives). “The model is almost gic orchestrates this process by creating perfect,” asserts Brown, because 100% internal workflows based on the integra- True Non-experimental True False Negatives “False Negatives” Positives Positives accuracy is not possible due to singletons tion of multiple functions: security/QC/ in the dataset. inventory control; predictive modeling; Fig. 2: The 50% mark is preferred by com- Using a computational strategy that identification of actives; and selection of panies because it produces a “manageable combines various statistical methods in- compounds to take forward into iterative workload.” cluding Bayesian algorithms, molecular screens. “Invest in building an SAR rather fingerprints, and chemical functionality, than amassing a lot of data points,” and we should be able to do all experiments Accelrys’s SciTegic Pipeline Pilot can HTS will pay off, advises Brown. in silico. But, at the moment, that is just a predict 74% of the active compounds in vision. The wet biology is still very impor- a large library while only screening 31% In Silico vs. Wet Lab tant,” particularly to validate hypotheses of the dataset. It does this by designing Because HTS data is not trustworthy for generated from virtual screens. iterative screens to build a structure- building predictive models, scientists These midstream course corrections activity relationship (SAR). are instead using the high-quality data and strategic shifts in designing and So why are companies still doing generated by biologists within project implementing high-throughput screens HTS “the wrong way,” asks Brown? Be- campaigns to design models, says Brown. are a natural outcome of the maturation cause they have the machinery built for They then rely on virtual screening to of HTS. Although the numbers are still sampling and screening entire libraries. identify well-defined compound sets, small, HTS has identified hits that have Without the appropriate infrastructure often a focused library, which can be used successfully made it into late-stage clini- and automation in place to cherry-pick to guide the preparation of HTS plates to cal trials and even to the marketplace. select compound sets efficiently and run screen against a particular target class. “Is that a sufficient pay-off at this point iterative screens, this approach could be Novartis’ Bojanic views HTS and in in time? Relative to the hype of the 90’s, too costly for many companies. silico screening as complementary tech- no,” says GSK’s Hertzberg. “Relative to Brown also contends that the problem niques. “The vision for the future is that what most realists thought, yes.” x

[24] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

IT/Workflow

SC08 Showcases HPC Muscle Systems IBM retains top spot Top 500 Trends grow, to ensure a cost-effective approach According to the latest Top500 list of for research. BlueArc , Isilon Systems, and on Top500 list. the world’s most powerful supercom- Panasas were among vendors presenting puters, IBM retained the top spot with solutions that deliver the capacity to store its Roadrunner system, deployed at Los massive amounts of life sciences data, BY SALVATORE SALAMONE Alamos National Laboratory. The system, tools to manage these data, and the per- he recent SC08 conference* marked a BladeCenter QS22 Cluster, had a peak formance to match data hungry CPUs in the 20th anniversary of the annual performance when running the Linpack today’s HPC systems. supercomputing conference and benchmark application of 1.105 petaflops. On the energy consumption front, T expo, with the application of high- (One petaflop is one quadrillion floating the Green500 list was released, which performance computing (HPC) in life sci- point operations per second.) ranks systems based on their energy ef- ences a major theme of the meeting. A close second was the Cray XT5 ficiency and processing power. Like the At Argonne National Laboratory, “We supercomputer at Oak Ridge National Top500 list, rapid change is the norm. award large blocks of time to accelerate Laboratory called Jag- The energy-efficiency of research,” said Peter Beckman, director of uar. The system, only the top supercomputers the Argonne Leadership Computing Fa- the second to break the increased by 17 percent cility (ALCF). The facility is home to one petaflop barrier, posted uad-core over the previous rank- of the world’s fastest computers for open a top performance of Q ings last June. science and is part of the U.S. Depart- 1.059 petaflops. processor Much of the discus- ment of Energy’s (DOE) effort to provide The new list reveals ‘ sion focused on the need leadership-class computing resources to that quad-core proces- based systems for organizations to use the scientific community. sor based systems have energy benchmarks to The main system at ALCF is the rapidly taken over the have rapidly get a better understand- Intrepid, a 40-rack IBM Blue Gene/P Top500. Already 336 taken over the ing of where electricity capable of a peak-performance of 557 systems are using them, is being used in the data Teraflops (557 trillion calculations per while 153 systems are Top500 at 336 center. Several panel- second). The ALCF won the SC08 High using dual-core proces- systems. ists noted that only one- Performance Computing Challenge sors and only four sys- third of the electricity Award, which compares systems using a tems still use single core used in data centers goes suite of benchmarks. processors. And seven to power the actual IT As part of the DOE’s annual Innova- systems use IBMs advanced Sony Play- equipment. Much of the energy is used for tive and Novel Computational Impact Station 3 processor with nine cores. cooling and a significant amount of power on Theory and Experiment (INCITE) One notable HPC debutante was is simply lost due to inefficient power dis- program, researchers are encouraged to Convey Computer Corp. It announced tribution within a data center. submit proposals to get time on ALCF the Convey HC-1, a hybrid computer that Daniel Reed, director of Scalable and systems. gets its processing power from a combina- Multicore Computing at Microsoft, noted Several life sciences projects were tion of Intel Xeon processors and Xilinx a once radical energy cost saving idea that selected in 2008. The awards included Field Programmable Gate Arrays. During is becoming more common today. “[HPC] 12 million processor hours for a com- one session, Rajesh Gupta ( University facilities do not have to be located next to putational protein structure prediction of California San Diego) discussed how researchers,” said Reed. In recent years, and protein design project and 5 million researchers using the Convey system and Microsoft, Google, and several other hours to investigate gating mechanisms of a UCSD proteomics application called In- organizations have moved their data membrane proteins. spect were able to accelerate their studies. centers to regions of the country where Other awards went to projects to sim- Using this combination, UCSD scientists the cost per kilowatt-hour of electricity is ulate and model synuclein-based protofi- can now perform unrestricted searches significantly below the national average. bril structures as a means of understand- of massive protein databases looking for With the ubiquity of high-speed Internet ing the molecular basis of Parkinson’s unanticipated modifications in peptide connections, providing researchers in disease, and large-scale simulations of samples. one location with access to an HPC cen- cardiac electric electricity. “Ten percent of ter in another is not out of the question. the machine can be used for discretionary Energy and Storage Center Stage That’s not a solution for everyone, but it projects,” said Beckman. Another theme was the importance of hammers home how HPC system energy storage systems and energy consumption, usage and costs have been elevated to the *SC08 – Austin, Texas. November 15-22, 2008 as HPC processing power continues to highest level of importance. x

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 25 ]

Enabling Technology. Leveraging Data. Transforming Medicine.

Featured Presentations by: KEYNOTE PRESENTATIONS BY: KEYNOTE PANEL: Chris Dagdigian, Founding The Future of Personal Genomics: Partner and Director of Technology, Platinum Sponsors: A special plenary panel discussion featuring... BioTeam, Inc. Robert C. Green, M.D., M.P.H., Professor of Neurology, Genetics and Epidemiology, Boston University School of Medicine and Public Health Clifford Reid, Ph.D., Chief Executive Offi cer, John Halamka, M.D., M.S., CIO, Harvard Medi- cal School Complete Genomics, Inc. Clifford Reid, Ph.D., Chief Executive Offi cer, Complete Genomics, Inc. Eric E. Schadt, Ph.D., Executive Dietrich Stephan, Ph.D., Co-founder and Chief Science Offi cer, Navigenics, Inc. Scientifi c Director, Genetics, Rosetta Inpharmatics/Merck Gold Sponsors: Research Labs

Event Features: $FFHVV$OO6L[7UDFNVIRU2QH3ULFH 1HWZRUNZLWK$WWHQGHHV +HDU7HFKQRORJ\DQG CONCURRENT TRACKS: 6FLHQWLÀF3UHVHQWDWLRQV Bronze Sponsors: Choose from Multiple Track 1: IT Infrastructure & Operations 3UHFRQIHUHQFH:RUNVKRSV Attend Bio-IT World’s Best Practices Awards Track 2: Informatics for Genomic Medicine Connect with Attendees Using CHI’s Intro-Net in silico Participate in the Poster Track 3: Predictive and Science Competition See the Winners of the following 2009 Awards: Track 4: Computational Chemistry %HQMDPLQ)UDQNOLQ Best of Show Track 5: Clinical and Medical Informatics Best Practices View Novel Technologies and NEW 6ROXWLRQVLQWKH([SDQVLYH Track 6: eHealth Solutions Corporate Support Sponsor: ([KLELW+DOO And Much More! Please mention keycode U02 when registering! Organized & Managed by: 2IÀFLDO3XEOLFDWLRQ Cambridge Healthtech Institute 250 First Avenue, Suite 300, Needham, MA 02494 3KRQH)D[ Toll-free in the U.S. 888-999-6288 www.Bio-ITWorldExpo.com

Ingrid Akerblom, Ph.D., Executive Director, Clinical IT, Rajarshi Guha, Ph.D., Visiting Assistant Professor, Thomas Paterson, Co-Founder & CIO, Entelos, Inc. Merck & Co., Inc. Informatics, Indiana University 6WHYHQ3HUULQ3K'&KLHI6FLHQWLÀF2IÀFHUALS Therapy John Apathy, Assistant Vice President, Business Systems Rishi Gupta, Ph.D., Senior Scientist, Computational Sciences Development Institute and Processes - Discovery, PreClinical, and Vaccines, Wyeth Center of Emphasis, 3À]HU Rudy Potenzone, Ph.D., WW Industry Technology Strategist %UXFH$URQRZ3K'3URIHVVRURI3HGLDWULFV6FLHQWLÀF Serge Guzy, Ph.D., Principal Scientist, Business for Pharmaceuticals, Microsoft Corporation Director, Center for Computational Medicine, Biomedical Development, XOMA David Powers, Lilly R&D Informatics, Cincinnati Children¹s Hospital Medical John Halamka, M.D., M.S., CIO, Harvard Medical School Center, University of Cincinnati James Reaney, Director, Research Markets, BlueArc Corporation 5HHFH+DUW6FLHQWLÀF0DQDJHU5HVHDUFK&RPSXWLQJGenentech Samuel (Sandy) Aronson, Executive Director of IT, Center for &OLIIRUG5HLG3K'&KLHI([HFXWLYH2IÀFHUComplete Personalized Genetic Medicine, Partners HealthCare, Inc. Peter V. Henstock, Ph.D., Statistics & Visualization, Target & Genomics, Inc. Mechanism Informatics, 3À]HU5HVHDUFK7HFKQRORJ\&HQWHU V.A. Shiva Ayyadurai, Ph.D., Executive Director, International Michael H. Roehrl, M.D., Ph.D., Pathology and Laboratory Center for Integrative Systems Research & MIT Faculty Karl Peace, Ph.D., Director of the Center for Biostatistics, College of Medicine, Massachusetts General Hospital Lecturer, M.I.T. Health and Human Sciences, Georgia Southern University David Ross, Ph.D., Director, Computational Biology, Celera Andreas Bender, Ph.D., Assistant Professor, Medicinal David Hodgson, Site Head, Research Informatics & John Russell, Executive Editor, Bio-IT World Global Head, Group Knowledge Exchange, Roche Chemistry Division, Leiden / Amsterdam Center for 3DXO5XWKHUIRUG&KLHI7HFKQRORJ\2IÀFHUIsilon Systems Drug Research Pharmaceuticals Alan Ruttenberg, Principal Scientist, Science Commons Brian Bissett, Staff Scientist, Molecular Properties, 3À]HU $QWRQ+RSÀQJHU3K'3URIHVVRUUniversity of New Mexico, Editor, Journal of Chemical Information and Modeling Johanna Sagemark, Head of Research Informatics, Structural Christoph Bock, Department of Computational Biology, Max Genomics Consortium, Karolinska Institute Planck Institute for Informatics Andrew L. Hopkins, D.Phil., FRSC, SULSA Research Professor, Translational Biology & Chair, Medicinal Informatics (ULF(6FKDGW3K'([HFXWLYH6FLHQWLÀF'LUHFWRU*HQHWLFV Mark S. Boguski, M.D., Ph.D., Founder, Resounding Health Division, University of Dundee Rosetta Inpharmatics/Merck Research Labs Incorporated Vibha Jawa, Ph.D., Senior Scientist, Clinical Immunology, Amgen Jon Schoonmaker, Chief, Technical Programs and Operations, 6WHSKDQ%RXU3K'&KLHI,QIRUPDWLRQ2IÀFHU'LUHFWRU SAFE-BioPharma Association Center for Biomedical Informatics, National Heart, Lung, Jeremy L. Jenkins, Ph.D., Research Investigator, Lead and Blood Institute (NHLBI) Discovery Informatics, Novartis Institutes for Raza Shaikh, Associate Director of Informatics, Chemical BioMedical Research Biology Platform, The Broad Institute Ken Buetow, Ph.D., Associate Director, Bioinformatics and Information Technology, National Cancer Institute Andrew Kaczorek, Eli Lilly and Company Angela Shen-Hsieh, President and CEO, Visual i|o Hugues Sicotte, Ph.D., Assistant Professor of Medical %DUU\$%XQLQ3K'&KLHI([HFXWLYH2IÀFHUDQG3UHVLGHQW $QGUHZ.DVDUVNLV3K'6FLHQWLÀF'LUHFWRU*HQHWLFVMerck Collaborative Drug Discovery, Inc. Research Laboratories Informatics, Biomedical Statistics and Informatics, Mayo Clinic Werner Ceusters, MD, Director, Ontology Research Group, NYS Dorothee Kern, Ph.D., Professor of Biochemistry, HHMI Ted Slater, Senior Manager, 3À]HU:RUOGZLGH7HFKQRORJLHV Center of Excellence in Bioinformatics & Life Sciences Investigator, Brandeis University Christopher Southan, Ph.D., EMBL European Joan Chambers, Senior Director of Marketing & Operations, Roy Kimura, Ph.D., Senior Scientist, Computer Assisted Drug Bioinformatics Institute (EBI) Publications, Cambridge Healthtech Institute (CHI) Design, Bristol-Myers Squibb John Speakman, Associate Director, Clinical Products and George Komatsoulis, Ph.D., Deputy Director, Center for Programs, Center for Bioinformatics and Information

CONFIRMED SPEAKERS Belinda Chen, Ph.D., Deputy Director, IDEAS, Institute for Information Industry Bioinformatics and Information Technology, National Technology, National Cancer Institute (NCI) Cancer Institute Chris Dagdigian, Founding Partner and Director of Donald R. Stanski, M.D., Vice President & Global Head, PH, Technology, BioTeam, Inc. Phillips Kuhl, Co-founder and President, Cambridge Modeling and Simulation, Novartis Pharmaceuticals Healthtech Institute Corporation Kevin Davies, Ph.D., Editor-in-Chief, Bio-IT World Ilya Kupershmidt, Co-founder and Vice President Products, Dietrich Stephan, Ph.D., Co-founder and Chief Science Michael Daoust, Google NextBio 2IÀFHUNavigenics, Inc. Giles Day, Senior Director BBC Informatics, 3À]HU Martin D. Leach, Ph.D., Executive Director, MRL IT Basic Susie Stephens, Ph.D., Principal Research Scientist, Eli Lilly & Co. Biotherapeutics & Bioinnovation Center Research & Biomarkers, Merck & Co. 0DUN7HSSHU3K'&KLHI([HFXWLYH2IÀFHUOptimata 'HUHN'HEH3K'6HQLRU*URXS/HDGHU6FLHQWLÀF Robert Lundsten, Director of Biorepository, Feinstein Informatics and Automation, Abbott Laboratories Alex Tropsha, Ph.D., Chair, Department of Medicinal Institute for Medical Research Chemistry and Natural Products, University of North David DeCaprio, Associate Platform Director for Chemical Subha Madhavan, Ph.D., Director, Clinical Research Informatics, Carolina School of Pharmacy Biology and Novel Therapeutics, The Broad Institute Lombardi Cancer Center, Georgetown University Matthew Trunnell, Manager, Research Computing, Broad John Dey, UNIX Operations Manager, Rosetta Robert Maguire, Senior Director MRL-IT Solutions Delivery, Institute of MIT and Harvard Inpharmatics, LLC MRL-IT Solutions Delivery, Merck Yaron Turpaz, Ph.D., Director of Integrative Computational Ron Dror, Ph.D., Senior Research Scientist, D. E. Shaw Research Bruce Martin, Vice President Software, Complete Sciences, Lilly Singapore Centre for Drug Discovery William Dyer, CRI, Principal Consultant, Bioinformatics, Genomics, Inc. Jonathan Usuka, Ph.D., Senior Director, Life Sciences National Cancer Institute Brian B. Masek, Ph.D., Product Manager, Tripos International Marketing, Accelrys Sujith Eramangalath, M.Sc, Senior Analyst & Team Lead, Paul Mattes, M.B.A., Director, Life Sciences, Health Solutions Ronald Waife, MPH, President, Waife & Associates, Inc. Healthcare, Frost & Sullivan Group, Microsoft Corporation Richard Ware, Senior Principal Business Partner, Regulatory -DFRE)DUPHU&KLHI7HFKQRORJ\2IÀFHUCambridge Computer Jomol Mathew, Ph.D., Director, Clinical Research Information Affairs, Global Drug Development Information Systems, Rick Franckowiak, Information Technology Director, Technology, Dana-Farber Cancer Institute AstraZeneca Pharmaceuticals Johnson & Johnson John McCarthy, Vice President, Product Management & Antony Williams, Ph.D. Founder, ChemZoo, Inc.; Host, Rich Furr, Head, Global Regulatory Affairs and Chief Strategy, Symyx Technologies, Inc. ChemSpider &RPSOLDQFH2IÀFHUSAFE-BioPharma Association John McGee, Manager, CI-Development, Renaissance Colin Williams, Product Manager, Biology & Bioinformatics; Brigitte Ganter, Ph.D., Senior Product Manager, Ingenuity Computing Institute, UNC Chapel Hill Product Management, PharmaChem, Thomson Reuters Systems 'DYLG0LOZDUG3K'&KLHI7HFKQRORJ\2IÀFHULinguamatics Mark van der Laan, Ph.D., Professor of Biostatistics and Meir Glick, Ph.D., Research Investigator II, Lead Discovery Peter Murray-Rust, Ph.D., Professor, Chemistry and Statistics, UC Berkeley Informatics, Novartis Institutes for BioMedical Research Molecular Informatics, University of Cambridge, Senior Tibor van Rooij, Bioinformatics Director, Génome Québec & Albert Gough, Ph.D., Vice President, Discovery Technology, Research Fellow, Churchill College Montreal Heart Institute Pharmacogenomics Centre Cellomics, Inc. Florian Nigsch, Ph.D., University of Cambridge and Jeffrey Wisemen, Ph.D., CTO and Vice President, Discovery, Robert C. Green, M.D., M.P.H. Professor of Neurology, Novartis Institutes for BioMedical Research Locus Pharmaceuticals Genetics and Epidemiology, Boston University School David Nunn, Ph.D., Principal Research Scientist, Matt Wood, Sequencing Informatics, Production Software, of Medicine and Public Health ChemInformatics, Wyeth Research The Wellcome Trust Sanger Institute Nigel Greene, Ph.D., Senior Principle Scientist, Jean-Christophe Olivo-Marin, Ph.D., Group Leader, Lei Xie, Ph.D., Principal Scientist, San Diego Supercomputer Toxicoinformatics, 3À]HU*OREDO5 ' Quantitative Image Analysis Unit, Pasteur Institute Center, University of California, San Diego Georges Grinstein, Ph.D., Professor & Director, Bioinformatics Dave Parrish, M.S., Director Informatics, Immune Program and the Center for Biomolecular and Medical Tolerance Network Informatics, University of Massachusetts Lowell www.Bio-ITWorldExpo.com

Clinical Research

2009 Predictions and Resolutions Waife and Sirabella see “on-shoring” as key trend for the New Year.

BY DEBORAH BORFITZ something, there will be high demand for he fundamental problems facing it and the price will rise. Playing games the clinical trials enterprise in 2009 with global macroeconomics is a bad idea. are the same ones that have been Plus, outsourcing overseas gives people Tplaguing it for years, industry in- another reason to dump on pharma these siders say, including lagging productivity days.” and overly complex study protocols that Sirabella agrees, adding that the impair patient recruitment and extend employee turnover rate in India is high, development timelines. Rather than further escalating costs. The situation in seek meaningful, long-term solutions, India is beginning to stabilize, as it is in companies tend to take shortsighted China, she adds. But there are many pock- corrective action to quickly sweeten the ets in the U.S.—Kentucky, Tennessee, and bottom line. Alabama among them—that are equally The nation’s ongoing economic tur- economical locations for setting up moil is also threatening the pharmaceuti- clinical research and data management cal industry with “unfamiliar challenges,” o short-term facilities. Companies could further save says Ronald Waife, president of Needham, N on overheads by having study managers Mass.-based clinical research consultancy financial advantage and other employees work remotely from Waife & Associates. Long discovery and ‘ home or other nearby location and only product cycles have traditionally shielded will ever be come on-site as needed. “Companies have pharmaceutical companies from the reali- permanent.” to become more comfortable managing ties of layoffs and facility shutdowns that employees that way.” Currently, it’s an ac- they’re now being forced to confront. Ronald Waife, Waife & Associates cepted practice only for study monitors. Odds are good that companies will be In partnership with ExecuPharm, looking at their targeted therapeutic areas has the recession evened out currency ex- Pfizer has successfully implemented the with a more discerning eye, says Lauri change rates, but Americans want to keep first extensive remote study management Sirabella, VP clinical strategic business jobs at home for reasons of patriotism, model—covering more than 300 em- development for Philadelphia, Penn.- convenience, and economic necessity. ployees in the U.S. and another 50 in the based recruiting company ExecuPharm. There is also tremendous pressure on the U.K.—by seeking people with “excellent Pfizer, which has its Lipitor patent expire new Obama administration to bring jobs communication skills who are proactive in the next few years, has decided not to back on shore. about preventing problems and handling pursue development of any other cardio- Furthermore, off-shoring can be an them when they arise,” says Sirabella, vascular drugs. Besides looking at their impractical course for reining in costs, who formerly headed up study and data core competencies, companies will need says Waife. “If you can’t manage people in management at Pfizer’s R&D facility in to consider which therapeutic areas will the next cubicle, why try to do it across the New London, Conn. “Most of what study have the greatest demand for treatments Pacific Ocean? Long-distance manage- managers do is electronically based today. based on worldwide population, age, and ment problems are enormous because of They don’t have to be in the building to disease trends. the lack of face time as well as language interact with people and manage a clini- Many companies will move into oncol- and culture barriers. The meaning of ‘yes,’ cal study.” Virtual employees also broaden ogy, which has strong potential growth, ‘deadline,’ ‘completeness,’ and ‘quality’ the talent pool that organizations can tap says Sirabella. There are also many all have [unique] cultural definitions. in less desirable geographic locations, promising leads to chase down, including All of that can be overcome with project including New London (Pfizer’s U.S. starving or treating tumors rather than management, but it’s hard to do ten time headquarters). removing them. zones away.” It’s unlikely that there will be any Off-shoring havens also quickly lose slowdown in the globalization of clinical On-Shoring their “salary advantage,” as is now being trials to Latin America, South America, In 2009, many companies may start proven in India. “No short-term finan- Asia, and Eastern Europe, says Sirabella, to consider on-shoring rather than off- cial advantage will ever be permanent,” largely because of the ready availability of shoring operations, says Waife. Not only says Waife. “If you find a cheap source of (CONTINUED ON PAGE 31)

[28] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

[GUEST COMMENTARY ]

Streamlining Regulatory eCTD Publishing Automated eCTD cross-references are applied, and all links ing documents for eCTD submissions: are verified. This process can delay a sub- • Document rendering—The better the submissions can save mission by weeks or months, eroding the quality, the less time required to review time before a patent expires and directly and update each document. companies time, money. impacting the bottom line. • Ease of Integration—Choose an application that employs stable, open protocols Building Submissions and standards (XML, Web services), BY SCOTT MACKEY The key to managing this tedious submis- which can fit into the infrastructure. When it comes to document and informa- sion process is to ensure that the docu- • Natural Workflow—The technology tion management, nowhere are stakes ment conversion and publishing tech- needs to fit seamlessly into the overall higher than in the pharmaceutical in- nologies used in building the submission document workflow of the organization. dustry. A day’s delay can mean the loss of can accurately and seamlessly present a Eliminate steep learning curves and the millions of dollars and, more importantly, high volume of information from multiple rate of adoption will soar. unnecessary postponements in getting a sources and create documentation suit- • Compliance Ready—Ensure that the ap- product approved. able for the final submission. plication has a successful history in life To accelerate and standardize the Robust PDF rendering is the backbone sciences/regulatory submissions. review process, the industry has been of many document management work- • Document Transformation—Applica- moving from antiquated paper-based flows and essential for the eCTD submis- tions should be able to handle all docu- submissions to an entirely electronic sion lifecycle. Improving rendi- system, leveraging XML and PDF tech- tions reduces quality assurance nologies. In fact, as of January 2008 the churn to only minutes per FDA mandates that all submissions are document (searching for miss- or a company that submits done using eCTD (electronic common ing or broken hyperlinks usu- F technical documents), an XML-based ally takes hours to complete). just 100 documents per structure containing high-quality PDF Manual document review and documents. Other global regulatory agen- editing can typically take two week, automating can save cies have also recognized the efficiencies hours per document; automat- 183 man-hours per week. of an electronic submission process. And ing document transformation even though paper submissions are still and generating submission- accepted in other international jurisdic- ready documents initially can reduce this ment transformation needs, including tions, thousands of submissions have time to ten minutes. For a company that document assembly, OCR text recogni- now been made electronically and the submits just 100 documents per week, tion, and document conversion. percentage of companies still using paper automating the process can save 183 man- • Scalability and Reliability—It’s impor- is dropping rapidly. hours per week or $572,000 per year tant to have a robust application that While the FDA introduced the eCTD (based on a salary of $60 per hour). The can scale to meet fluctuations in vol- standard to simplify and accelerate the cost savings, projected over submissions umes coupled with the critical nature of submission process, organizations still running into the hundreds of thousands speed to market. struggle to automatically and accurately or even millions of pages, have proven to A robust regulatory publishing infra- convert submission documents from vari- be enormous. structure is designed to be a foundation ous formats into submission-ready PDFs. Submission-ready PDFs are flexible for growth for life sciences organizations. Key documentation is often compiled and contain an almost limitless set of This critical infrastructure component and maintained on a series of unmanaged features, including elements like access takes advantage of existing investments Microsoft Word documents and even control, standardized hyperlink styles, in skills and technology to achieve docu- hand-written forms. The flood of docu- embedded fonts, comprehensive tables ment management efficiencies that sup- ments that flows through the submission of contents, and bookmark structures. port the critical work being conducted on process often lacks conformance to any These PDF documents are also scalable behalf of the population—and corporate common standard. To ensure the rapid throughout the submission lifecycle, shareholders. The right eCTD technolo- and successful review of a submission, particularly important given the FDA’s gies can result in significant cost-savings, these documents must go through an ex- increasingly assertive use of post-mar- rapid ROI, and faster time to market. x tensive manual Quality Assurance process keting “risk minimization action plans” to ensure they have the correct bookmark (RiskMAPs). Scott Mackey is the director of product structures, fonts are properly embedded, Here are a few key steps that organiza- management for Adlib Software. He can be hyperlink styles meet FDA guidelines, tions should keep in mind when compil- reached at [email protected].

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 29 ]

Clinical Research [GUEST COMMENTARY ]

Three Key Ingredients to Cost-Effective Clinical System Interoperability CDISC ODM, Web-services, and the right business model smooths clinical trials.

BY RON RUBINSTEIN cal trial sponsors can eliminate redundant fit with ODM), security models, and a Not long ago, integrating clinical trial sys- tasks, increase data quality, and accelerate common Application Program Interface tems was a tricky and expensive business. the flow of critical information, giving (API) to link loosely-coupled, disparate There was no set standard for storing data, them a distinct competitive advantage. systems. The ubiquity of Web technology or defining how—and in which format—it and related development tools has dra- would be shared between systems. Every The First Ingredient: CDISC ODM matically reduced the time, complexity, data process was specially constructed In an attempt to reduce these integration and cost of integrations using this meth- on a by-project basis and because of this, costs, the clinical trial industry has pushed od. Because of the synchronous, direct, much time was spent planning and de- for data standards that would take some and API-based nature of data transport signing the integrations. Implementation of the guess-work and proprietary nature using Web services, information-sharing risks and delays were the norm, and sig- out of how data is modeled and format- is real-time and specific data can be nificant costs were incurred from having ted when stored and transported. For requested on the fly—without the need to constantly custom-build, test, validate, example, the Clinical Data Interchange for pre-configuration or consideration and maintain the integrations. Standards Consortium (CDISC) has taken of how data is generated. Web services As some vendors became more expe- a lead role in defining data standards for adoption is the driving force behind IT rienced with custom integrations, they clinical research technology—notably, the architecture “movements” such as Service were able to reuse some elements of previ- Operational Data Model (ODM). Oriented Architecture (SOA), however ous integrations, thereby reducing costs ODM defines a standard XML file for- the cost savings and benefits of Web ser- and implementation risks to a certain mat for the transport of clinical data and vices integrations are well-documented extent. But, even this improvement was metadata between clinical systems. Re- and independent of SOA. confined to instances in which the recent reports show an impressive adoption With ODM adoption reaching a point quirements were similar between studies. rate of CDISC ODM and by becoming of critical mass, and Web services integra- In order to fully achieve clinical system the de facto standard, it is fast becoming tion gaining momentum across the indus- interoperability, while reducing costs and pharma and biotech companies’ format of try, one would expect to see Web-based complexity, sponsors need to use three key choice for clinical data integrations. This data sharing between clinical systems ingredients: CDISC ODM, Web services, has, in turn, compelled vendors to create take a more prominent role. and the right business model. ODM-compliant data interfaces with the expectation that other systems will also Mixing in the Right Business Model Why Integrate? adhere to that standard. The final important ingredient to en- Stepping back a moment, it’s important suring the proliferation of clinical trial to remember why clinical trial system Blending ODM with Web Services system integration, along with the realiza- integration (while painful, at times) is Data format standards like ODM take tion of true interoperability, is the adop- important in the first place. The typical some of the pain and risk out of inte- tion of the appropriate business model by pharmaceutical R&D department uses a grations, but are not a replacement for data service providers in the industry— multitude of different software systems to designing the data transport mechanism. specifically, in the case of electronic data automate its work. These systems are gen- This mechanism must be configured to capture and clinical data management erally used in isolation, typically known accommodate system details, the data systems. as a “siloed” approach, and provide ben- that is to be shared, and other elements Business models vary by vendor when efits for specific aspects of a large-scale specific to the study at hand. In many it comes to structuring the costs of doing clinical study—yet they also often affect cases, this process cannot be circum- business with clinical sponsors and common processes and use overlapping vented, but many IT and software de- service providers. Interestingly, some of information. velopment organizations have adopted these models actually deter sponsors from As the number of clinical systems and a low-cost, standards-based approach trying to create robust interoperability the demand for efficiency increases, there of Web services to integrate between architectures because “excessive” data- is a growing need to share this data. And, systems. sharing translates into high service fees. by cost-effectively integrating technolo- Web services introduce data shar- Conversely, true support for widespread gies and automating data-sharing, clini- ing standards using XML (an excellent interoperability will most likely drive a

[30] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

looking for.” That doesn’t mean they can Predictions be cavalier about operations. “Operational (CONTINUED FROM PAGE 28) inefficiency is rampant and wastes a lot of trial volunteers who don’t have good ac- money and time and that, too, has an im- cess to medical care. The sheer volume of pact on sales and shareholder value.” oncology trials has also made recruitment A second ideal New Year’s resolution quotas hard to fill in the U.S. would be for companies to be more intro- Outsourcing will continue unabated spective. Why do they ask so many ques- for the next five to ten years, predicts Sira- tions during a clinical trial that don’t have bella, because it is difficult for companies relevance to proving or disproving the to own staff in economically volatile times. endpoint? Why does a particular work- There is plenty of competent talent for flow process have seven steps rather than sale, as key leaders and much of the staff two? And why are there multiple cycles of at contract research organizations and quality checking in every department? other vendors are former pharmaceutical Forward-thinking companies will company employees. hey [study probably want to start paying attention T to the social networking movement, says Suggested “Resolutions” managers] don’t have Waife, pointing to President Obama’s If companies resolve to do anything this ‘ fund-raising efforts due to his virtual pres- year, ending internal “partisanship” would to be in the building ence on MySpace and Facebook. The “an- be a good place to start, says Waife. No- to... manage a clinical archic principles” of social networking are tions about who is right and who is wrong not a good fit with the scientific method, can take on almost religious proportions study.” but these sites may well have applicability at some organizations. “Between depart- in the notoriously knotty area of subject Lauri Sirabella, ExecuPharm ments there is [mutual] mistrust and recruitment. even disrespect.” If lasting change is going A “radically different” clinical trial to happen, it may first be necessary to re- circumstances such as today’s financial, management system (CTMS) would also organize the power grid to accommodate operational, and public perception prob- better suit the industry than the current, cooperation. lems, and have little training in dealing “impossibly over-featured” spoke-and- “You can’t really work together if you with difficult decision making. “[They] wheel platforms for exchanging informa- have fixed ideas,” says Waife. Heroic sto- don’t know what to do about it. They want tion that serve no user type particularly ries about how everyone pulls together it to go away or want people to make nice well, says Waife. Like it or not, pharma re- to get a submission out to the Food and with each other.” search happens in silos and each of those Drug Administration at the last minute Inexperienced leaders look for quick disciplines “deserves a tool that serves conveniently overlook the inefficiency that fixes, which Waife views as a sign of their their operational excellence.” prompted such valiant action. “Heroes are boredom with operations. “It may be legit- It’s not that integration isn’t important, formed in battle and the battle goes on.” imate that the most important thing they adds Sirabella. One of the most helpful Weak executive leadership, fearful of can do is acquire a product for the pipeline features of a CTMS, which surprisingly losing talent, only adds to the partisan- that is nearer to market [than anything few systems have today, is the ability of its ship problem, says Waife. Pharma execu- in the existing portfolio], which results component databases to share informa- tives have not had to experience adverse in sales and profit their shareholders are tion with each other. x

trend to reduce or ultimately abandon Making Interoperability a Reality business efficiencies to materialize. Ven- data-transport fees where those still exist The adoption of CDISC standards, Web- dors that offer cost structures supportive (similar to developments in the telecom services, and the right business model will of data interchange, in addition to stan- industry over the past two decades). Some be the gateway to providing pharma and dards-based approaches that empower examples include charging per data-point biotech companies with the necessary partners to develop joint solutions, will and assessing a service fee per every data tools to reduce the complexity of inte- help create those market conditions and extraction. In addition, very few Web ser- grating clinical systems. The end result bring the vision of true clinical interoper- vices vendors encourage third parties to is seamless data sharing, along with time ability to fruition. x independently exploit the API to develop and cost reductions in the implementation their own integration solutions, as they and use of IT for research processes. Ron Rubinstein is senior director of product prefer to offer such integrations exclu- As always, the right technology must marketing, Medidata Solutions Worldwide. sively as in-house services. align with market conditions in order for Email him at [email protected].

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 31 ]

Economic Development

BioQatar: A Splash in the Desert The tiny Arabian principality is luring scientific entrepreneurs, venture capitalists, and American universities to spur biotech development.

BY KEVIN DAVIES OHA, Qatar—It is not every day that one attends a conference where the housekeeping an- Dnouncements include the location of the prayer room or the moderator is a celebrity television news anchor with Al Jazeera. But it is not every day that a tiny Arab principality decides to make a concerted effort to woo foreign biotech and capital investment. Judging by the delegates and proceedings at the BioQa- tar 2008 symposium*, Qatar is well on its way. Qatar is a tiny nation nestled between Kuwait to the north and the United Arab Emirates to the south. The 200,000 na- tive Qataris are outnumbered 3:1 by for- Christopher Evans considers moving part of his operation to Qatar. eign workers from countries such as India and the Philippines. While the newly gies, with lectures regularly streamed live country. The two major health problems opened Museum of Islamic Art, designed from Ithaca to Doha. In all, Doha’s Educa- facing Qatar are cancer and diabetes. by I.M. Pei and sitting majestically on its tion City boasts 70 nationalities among its The BioQatar symposium attracted an own island in the Doha Corniche, is cer- faculty and students. impressive list of speakers, including out- tain to draw throngs of visitors, the Qatar Saoud said a priority was to build a going NIH director Elias Zerhouni, Nobel Foundation (QF) has its sights on modern reputation for research excellence and laureate and stem cell pioneer Sir Martin science and medicine rather than antiqui- culture. “Education is not enough on its Evans, and one of the U.K.’s best known ties. “Our goal is to become the scientific own,” he said. “We must become genera- venture capitalists, Sir Christopher Evans center of excellence in the Middle East, tors of knowledge, not just consumers of (no relation). Zerhouni described the shift and rekindle the reputation for learning it.” To that end, the Emir of Qatar has from acute to chronic diseases; the impact in the region from several centuries ago,” dedicated 2.8% of the nation’s GDP to re- of an aging population on the frequency said QF president Fathy Saoud. search and technology. Given that Qatar of neurodegenerative diseases; the threat The country has already fostered is the world’s third largest producer of of re-emerging infectious diseases; and close ties with several U.S. universities, liquefied natural gas, Saoud was being the new paradigm of the 4 P’s—predictive, including Georgetown, Carnegie Mel- a tad modest when he noted that was “a personalized, participatory, and preemp- lon, and Texas A&M. Last May, the Weil very credible proportion by international tive medicine. Cornell Medical College in Qatar, on the and regional standards.” Speaking in a thick Welsh accent, sand-swept outskirts of Doha, proudly Christopher Evans pointed out that saw its first class of 15 medical students Three Pillars “medical science literally started here— take the Hippocratic Oath before starting Qatar is seeking to attract investors in Islamic medical science was fantastic their residencies in the U.S. Cornell is the fundamental and applied research, and 1000 years ago.” The first vaccine and the first American university to offer its MD technology development, focusing on first pharmacies were developed in the degree overseas, with the same admission three areas: energy, the environment and Middle East, he said, but “Islamic medi- standards and curriculum as the parent biomedical research. There was talk of cal science got lost somewhere in the past institution. The Qatar school has enrolled advancing personalized and predictive 1000 years,” while Europe moved out of students from more than 30 countries, health care, bringing academic health the Dark Ages. and boasts enviable e-learning technolo- sciences closer to hospital practice and Whether Evans’ firm Excalibur (a suc- *BioQatar 2008 symposium, Doha, Qatar; 17-18 Novem- delivery at the bedside and raising the cessor to Merlin Ventures) will charge to ber 2008. standard of health care throughout the the rescue remains to be seen, although

[32] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

Save an extra $300.00 off the current rate!

his post-conference private On the Move audience with Her Highness One scientist-entrepreneur Sheikha Mozah signals genu- who is definitely moving to ine interest. Evans says he is Qatar is John Hassard, co- excited about building one or founder of deltaDOT . He’s two local biotech companies launching a subsidiary called February 23 – 24, 2009 The Sheraton Carlsbad Resort & Spa Carlsbad, CA in Qatar and stimulating US QSP Proteomics, which will and European investment. work on therapeutic/vac- Is Your His resume speaks vol- cine development and new Featured Keynote umes: the venture fund has technologies. Speakers: Organization built 45 medical companies, Niels Porksen talked about Prepared for developed 200 novel med- Lilly’s move to establish new the Promise of icines, and a net worth of centers outside the U.S., in- Madhulika Agarwal, MD, MPH Health 2.0? some $7.5 billion. “Our firms cluding Singapore, India, Chief Patient Care Services Officer have saved many lives,” he China, and possibly Qatar, Veterans Health Administration (VHA) • Social Networks said. Evans said Qatar’s future where there may be greater looked fantastic, but it “must local expertise. “We’re will- Kate Christensen, MD Medical Director • Patient-Provider Portals cover the whole waterfront— ing to do less in Indianapolis Internet Services Group don’t specialize in cancer or and more in the rest of the Kaiser Permanente diabetes.” He praised the facil- world….. It’s a way to get us • Personal Health Records ities currently under construc- into the future,” he said. Karen Kocher and Platforms tion, and said it was important Phillipe Froguel, a ge- Chief Learning Officer for Qatari students to travel neticist working at Impe- CIGNA • Gaming Interfaces before returning home to es- rial College, London, said James R. Mault, MD, FACS tablish businesses. that diabetes offered some of Director, New Products and Business Qatar’s health minister, the best examples of genomic Development Health Solutions Group • Mobile Applications Ghalia Al Thani, discussed the medicine success, such as the Microsoft Corporation country’s priorities: diabetes high cure rate of neonatal dia- • Privacy and Security is rampant, the median age is betes. Froguel is conducting Deven McGraw 76 and improving, but a major a genome-wide association Director, Health Privacy Project Center for Democracy & Technology (CDT) cause of premature death is study in Morocco to compare • Interoperability something more mundane: the susceptibility of an Afri- Jody Pettit, MD traffic accidents (299 in 2006) can/Arabian population with Strategic Leader, PHR Work Group • Health Information Search and workplace injuries. Con- the genes identified so far in Certification Commission for Healthcare sanguineous marriages (54%) Europeans. Information Technology (CCHIT) Former • TeleHealth and Remote were being addressed from Despite the enormous op- Health IT Coordinator Monitoring social and medical aspects. portunities Qatar presents, Office for Oregon Health Policy & Research The director of Qatar’s Sci- manpower is severely lim- Former Project Director ence and Technology Park, ited, and an entire biomedi- Oregon Health Information Security & • New Media Marketing Eulian Roberts, said that the cal infrastructure must be Privacy Collaboration country’s flagship teaching built, including regulatory hospital, the 380-bed Sidra policies, animal facilities, and Ken M. Shachmut Medical and Research Center, safety standards. And there Senior Vice President Strategic Initiatives, SAVE AN EXTRA Health Initiatives and Reengineering $300.00 OFF THE will be finished by 2011. “Co- was little sign of collabora- Safeway, Inc. CURRENT RATE! operation is essential to help tion with other Arab nations, To qualify, please contact us at Qatar become recognized as which could prove a fatal 800-767-9499 with a hub for research, innova- mistake. promotional code TXM424* by Organized By: tion and entrepreneurship,” Evans sees strong paral- January 30, 2009. he said. The Science Park is lels between Qatar today and *Government and workshop a truly 21st century conglom- the U.K. in the early 1990s, only rate not included eration of new construction when biomedical companies surrounding a massive con- began to make a splash in the vention center, seeking to lure markets. “I’m very pro Qatar,” To register, please visit www.worldcongress.com/connectivity biotech companies to join the said Evans. “Think selfishly • Phone 800-767-9499 • Fax: 781-939-2543 likes of Shell and Conoco- for five years and globally • Email: [email protected] Phillips as the first tenants. thereafter.” x

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 33 ]

New Products

Enriched PCR RainDance Technologies introduced its Sequence Enrichment Solution for targeting sequencing applications of the human genome. RainDance specializes in microdroplet-based solutions for human health and disease research. The new RDT 1000, part of the Sequence Enrich- ment Solution, generates picoliter volume discrete PCR reactions in droplets at the unprecedented rate of 10 million per hour. The RDT 1000 generates replicate reactions to ensure confidence in sequencing results, and the “open design” easily integrates with all next-generation sequencing platforms. RainDance is also offering customer-specific DNA primer cost and requiring only a few micrograms of sample DNA. The libraries to enable the amplifica- solution is currently in use at the Genome Sequencing Center tion of hundreds to thousands of at Washington University and the J. Craig Venter Institute. It genomic loci in a single tube. The will be commercially available in the first quarter of 2009. complete solution minimizes errors Product: Sequence Enrichment Solution and maximizes the efficiency of PCR Company: RainDance Technologies by leveraging sensitivity and speci- Available: Beginning of 2009 ficity, while reducing study time and For more information: www.raindancetech.com

Security Speed In order to speed rollout of Lawson Security 9.0, Lawson Security FastTrack offers pre-built content to help reduce the time needed to define, implement, and test the security solution. FastTrack includes a preconfigured and customizable set of task-based security classes, processes, templates, training modules, and tools to set up Lawson Security 9.0 in as little as six weeks. The solutions was used by Moffitt Cancer Center, and provided more than 75% of the institution’s security requirements out of the box. Product: FastTrack Company: Lawson Security Available: Now For more information: www.lawson.com Compound Registry Symyx has released Registration 1.5, replacing ISIS- based Symyx ChemBio AE software. Registration 1.5 Mesoscale Simulations offers R&D organizations of all sizes a flexible, fully- featured, Isentris-based approach to building and man- Materials Studio 4.4 is the newest modeling and simulation plat- aging consistent, searchable, and sharable corporate form from Accelrys. Studio 4.4 includes Mesocite, a new module compound registries based on Symyx’s chemical rep- that enables the simulation of particles incorporating ring struc- resentation technology. Registration 1.5 automatically tures, electrostatics, and allows users to incorporate more me- verifies drawing conventions, normalizes structure data, soscale simulations than previously available. It also allows for an and checks for uniqueness before assigning corporate intermediate-level of simulation between the fine-grained atomis- identification numbers and storing substances in the tic models and the more coarse-grained Mesotek models. compound registry. Product: Materials Studio 4.4 Product: Registration 1.5 Company: Accelrys Company: Symyx Available: Now Available: Now For more information: http://accelrys.com For more information: www.symyx.com/isistransition

[34] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

Lab Automation, LIMS and Data Integration Signature Series The Right LIMS Delivery Method

By J. Thomas Kent other costs (testing, customization, updating security, confidentiality, and brings excellent ntil recently, a commercial Laboratory and maintenance). The major difference is economies of scale, since the vendor does Information Management System a staged payment for the software. not need a massive investment in hardware (LIMS) was beyond the reach of Hosted LIMS is a commercial LIMS and services to provide the latest version Umost small and medium-sized labs. provided to customers in a different way. of the product. For this reason, true SaaS So labs often resort to “home-grown” solu- Rather than requiring the customer to pur- vendors can often sell services at a lower tions to save costs—only to find they’ve chase hardware and software, the customer cost than traditional vendors. And SaaS spent 2-10 times as much as a comparable simply uses the software that is running at offerings often embrace the idea of an commercial LIMS would cost. A number the vendor’s site. The vendor will have built online community where users can share COMMENTARY of commer- a front-end to the product that allows the best practices, collaborate, and easily cial LIMS lab to connect to the software through the purchase third-party software to meet their vendors are now offering “hosted”, “rented”, Internet and run the product. In general, emerging needs. “SaaS-based”, or “in-the-cloud” LIMS aimed hosted software providers do not rewrite directly at small to medium-sized labs. While their products to take advantage of new, What’s Best for Small Labs the vendors make these products sound Internet-based technologies, but simply There is no one-size-fits-all approach to LIMS, appealing, it is often difficult to read between put a different front-end onto what is often but budgetary and resource pressures often the lines of vendor advertisements to deter- a very old technology base. In addition, the preclude purchasing or renting a LIMS solu- mine the precise features, functionality and provider often must create a unique instance tion. Labs are looking for vendors who will benefits—as well as the differences among of the underlying product and infrastruc- deliver robust LIMS solutions in a less expen- the various offerings. ture for each user organization. This can sive manner. Today, we are seeing several represent either a slight increase in cost LIMS vendors who provide solutions based Many Options to the vendor, or a financial nightmare for on either the hosted or SaaS approaches Purchased commercial LIMS are, in customers. Each instance of the software defined above (including Sciformatix, Bika, general, large, complex, feature-rich requires its own hardware infrastructure, and LabLynx, and Promium). products that were designed to be sold the services burden is even greater. Imagine The advantage of such LIMS is that, if to large labs for in-house deployment. trying to upgrade dozens or even hundreds of designed properly, they can provide superb The lab purchases licenses for the LIMS instances of software and hardware without capabilities at a fraction of the cost of pur- software, and also buys the servers, causing massive user disruption and trying chased or rented systems. And they can, if storage devices, peripherals, and other to synchronize versions for all the different built on new technologies, provide improved software licenses (such as databases, customers. ease of use, robust reporting and compliance load balancers, backup software, etc.). “In-the-cloud”/Software-as-a-Service capabilities, 24x7 availability, easy upgrades, The lab installs the software in a test envi- (SaaS) is a relatively new software delivery and access to a community of users who ronment, performs a thorough evaluation, model for deploying business services on- can enhance the LIMS over time. and often pays hefty fees for the vendor demand, in a cost-effective manner. While Whenever a new idea such as SaaS or another professional services group to it bears some resemblance to a hosted arises, many traditional suppliers adopt the perform customization. Once in produc- model, there are very specific requirements. terminology without making the necessary tion, the software is updated and patched, True SaaS-based offerings are usually built changes in products, services, or delivery and hardware is maintained, as needed. from the ground up to take advantage of mechanisms. It is very difficult to retrofit Most small and medium-sized labs cannot service-oriented architecture, Web 2.0, SaaS into a product that was built ten or afford this expense. and other relevant technologies that are 20 years ago. It takes domain expertise, Rented commercial LIMS represents an designed to deliver sophisticated capabili- technology skills, and time. Lab managers effort by LIMS vendors to appeal to small ties on-demand. SaaS offerings are built for should be careful to separate fact from and medium-sized labs. A rental approach multi-tenancy, where only a single instance marketing hype in making such a critical is almost identical to the above purchased of the application is required, with a single decision. approach, except that the lab rents the soft- database that can manage the data for ware rather than purchasing the license out- multiple customers, while keeping customer J. Thomas Kent is President and Founder of Scifor- right. All the other purchases—hardware and instance data separate from all other cus- matix, a provider of SaaS-based LIMS solutions. He additional software—remain the same, as do tomers. Done correctly, this model provides can be reached at [email protected].

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 35 ]

SPECIAL ADVERTISING SECTION

Lab Automation, LIMS and Data Integration Signature Series

SYMYX Operational agility and productivity with an enterprise Notebook

ecent industry reports indicate that many pharmaceutical and biotech organizations are in the market for a single, enterprise electronic laboratory notebook. And with good reason. A single Rnotebook fosters workflow optimization, process consistency, collaboration, and improved productivity through knowledge sharing. A single notebook deployed across the enterprise also solves many of the information challenges within R&D, including data silos, information access bottlenecks, inconsistent data, and high costs of ownership.

With the acquisition of MDL, IntelliChem, and Synthematix, Symyx’s experience in delivering scientific Notebooks extends over 15 years with some of our earliest notebooks delivered using ISIS technology (formerly owned by MDL). In the last few years there has been a shift in our customers’ notebook goals. Where individual disciplines previously had independent notebook strategies, these are now consolidating into a single, multi-disciplinary, enterprise notebook approach. Benefits realized by a unified notebook include:

• Consistent and comparable processes • Configurability to support the diverse An enterprise electronic lab notebook • Centralized and accessible data needs of different domains without requir- streamlines the capture of chemical and • Improved data consistency ing extensive customizations with long- biological information in regulated and non- • Agility for scientists to move between term cost of ownership implications regulated environments. projects • Ability to deploy in regulated and non- • Lower cost of ownership regulated environments without requir- • Streamlined regulatory compliance ing non-regulated departments to bear from the notebook makes it a priority to also regulated compliance burdens integrate best-of-breed tools, information, Furthermore, investments in global • Straightforward extensibility of core func- and software into the notebook. enterprises and contract research organi- tionality using modern architectures and zations are calling for added support of a programming languages that foster easy Accessing content in context single notebook that can be used world- integration The electronic notebook changes the way wide across business ventures. Working • Global deployment with a centralized data scientists interact with information. Not only with customers, Symyx has been quick to repository and infrastructure where all data is the data captured by the notebook fully respond to the new strategies, releasing can be readily accessed and shared searchable within and across experiments Symyx Notebook in 2008 to address the but third-party information is automatically following core challenges faced by traditional With the deployment of our enterprise built into the scientist’s workflow. Without electronic notebooks: notebook, Symyx is seeing a shift in the leaving their notebook, for example, scien- way scientists interact with their traditional tists can look for commercial compounds • Customizable workflows complimented desktop software and laboratory equipment. or known synthesis pathways and popu- with fine-grained security enabling indi- The notebook is becoming the hub of the late the experiment details. As scientists vidual parts of an experiment to be carried scientist’s day, the place where one orches- enter molecules they are automatically out by many individuals across different trates the design, execution, analysis, and informed about in-house and commercial organizations reporting of experiments. Driving research availability.

[36] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

SPECIAL ADVERTISING SECTION

Integrating with other workflow/ lab automation solutions The electronic notebook is ideally positioned to facilitate integration with existing workflows, apparatus, software, and colleagues. Common integration points include:

• Mettler-Toledo and Sartorius balances, enabling scientists to retrieve calibration information and record actual amounts directly from balances • Chromatography Data Software (CDS) such as Waters Empower • Data repositories such as Waters NuGen- esis Scientific Data Management System (SDMS) • Microsoft Excel spreadsheets • Statistical analysis software such as SAS JMP An enterprise electronic lab notebook fosters improved collaboration and decision making • Kinetic modeling software such as Phar- within and across project teams. sight WinNonlin

Enhancing the value of new level, Symyx Isentris® decision support notebook information software enables scientists to search over Sir Isaac Newton said “If I have seen farther both in-house data and commercially avail- than others, it is because I was standing able data. For example, the Symyx reaction “Customers are realizing on the shoulders of giants.” The ability to databases provide access to over 7 million use and re-use the knowledge of others is chemical reactions. These reactions can be significant efficiency critical in improving experimental success. combined with in-house reactions captured improvements and cost A key focus of the Symyx notebook strategy through the notebook and accessed through is addressing existing notebook shortfalls a single search in Isentris. reductions by cloning for accessing notebook data. As well as supporting full text searching within and To summarize, the electronic notebook experiments, using pre- across experiments, including searching is entering a new era. R&D organizations defined data entry forms, image annotations, it is also important to are looking for improved productivity and offer scientists extensive molecule and greater operational agility—the ability to do and storing frequently reaction searching techniques that include more with less. Today’s enterprise notebook used procedures the ability to search reaction centers for strategies are raising the bar for notebook transformations. vendors, driving us to deliver software that electronically. Scientists streamlines scientific workflows and fully Chemistry excellence leverages our customers’ valuable intel- using our notebook A key request from customers has been to lectual property across disciplines, depart- have estimated a 20% integrate Symyx chemistry into the notebook, ments, and collaborating organizations. enabling both scalability and comprehensive increase in productivity searching. Symyx chemistry is the most widely adopted and proven chemistry in simply by streamlining the industry. Symyx’s online DiscoveryGate data input.” application is powered by Symyx chemistry enabling thousands of researchers to query — Cheryl Lund Ph.D., commercial bioavailability, synthesis, and analytical notebook sourcing databases containing over 30 product manager at Symyx million molecules and 17 million reactions. For more information on Symyx notebooks, Taking the value of notebook information to a visit www.symyx.com/notebook6

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 37 ]

SPECIAL ADVERTISING SECTION

Lab Automation, LIMS and Data Integration Signature Series

GENOLOGICS LIMS to Enable Translational Research

Q: What are the major challenges faced space are fairly different from laboratories in in translational research? other domains. LIMS have been utilized for JD: Translational research holds huge prom- process chemical, petrochemical, food and ise to drive the advancement of molecular- beverage, materials testing, and many other based diagnostics, and enable early disease procedural oriented areas for many years. detection and personalized medicine. But, But it’s only been since roughly 2000 that the very essence of translational research, life sciences research laboratories have being highly collaborative, multi-disciplinary, been at sufficient scale to require LIMS translating insights and results from one solutions. domain to another, creates many “systems” LIMS for translational research require type challenges for research organizations. something quite unique and different than Many translational research efforts are traditional LIMS. A successful life sciences taking scientific discoveries and applying research LIMS requires a far higher degree them to a clinical context, for example, of flexibility, configurability, deep instrument biomarker validation studies. This means integration, science data integration, bioin- higher specimen counts versus typical dis- formatics tools and algorithm automation, covery experiments. It means clinical data collaboration, and open data access. Only must be integrated from many sources and by focusing exclusively on life sciences and marshaled into the research domain. It often James DeGreef translational research, by providing a highly involves many collection sites, bioreposito- flexible, configurable and usable LIMS plat- ries, core technology labs and instruments Vice President Market Strategy form, by forming tight business and technical generating large data sets across sciences GenoLogics Life Sciences Software relationships with leading instrumentation and institutions. providers and by building seamless integra- For example, it’s becoming more com- and figure out the bounds of Translational tions to genomics, proteomics, and other mon for large pharmaceutical companies to Research Informatics (TRI), as it fills the instrumentation, can a LIMS provide real collaborate with academic medical centers gap between patient care and clinical trial value for life scientists. for specimen, expertise and technology systems, and what many are now calling access, as well as for institutes working translational bioinformatics. Translational Q: What are the trends for data integra- together in a highly collaborative manner Research Informatics is the integrated tion within translational research? such as under the CTSA program. All the end-to-end solution that provides: study JD: To flourish, translational research people, expertise, specimens and instru- management; electronic patient question- requires a knowledge-driven ecosystem, in ments are available, but there are enormous naires; clinical information aggregation for which constituents generate, contribute, logistical challenges of study and specimen research purposes; federated biorepository manage and analyze data available from management across an enterprise and chal- management; LIMS purposed for analytical all parts of the landscape. The goal is a lenges with data integration – both clinical and core technology labs; genomics, pro- continuous feedback loop to accelerate the and research data – and integrating data teomics and other sciences data manage- translation of data into knowledge. Collabo- from different technology systems. Really, ment; and traceability, data integration and ration, data sharing, data integration and the missing component is integrative trans- marshalling for analytics. standards are integral. Only by seamlessly lational research informatics. structuring and integrating these data types Q: What are the trends for LIMS within will the complex and underlying causes and Q: So how would you define translational translational research? outcomes of diseases be revealed, and research informatics? JD: The requirements for LIMS in the effective prevention, early detection and JD: It took GenoLogics a while to define translational research and systems biology personalized treatments be realized.

[38] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

SPECIAL ADVERTISING SECTION

Solutions for Translational Research Informatics

GenoLogics believes informatics is the key enabler for translational research. Many of our customers have brought challenging requirements to us and overall we have been successful in providing valuable solutions for translational research informatics. We work very closely with our customers, both with pharmaceutical and biotech companies and with academic medical centers and research universities. Many of our customers’ informatics groups consider us to be an extension of their team, and we work together More specifically, translational research Sciences Center-Houston that have imple- on an integrated solution to solve requires investigators to have ready access mented our research informatics platform their challenges and provide value to two critical types of data: (1) clinical infor- to manage multiple genomics, proteomics, to principal investigators, study mation, including data contained in hospital and other core technology laboratories. The and lab managers, technicians systems and medical records, pathology primary focus has been on integrated data and biostatisticians. reports and diagnostic labs, clinical trials management to provision analytics, but also GenoLogics is taking an inte- systems, study participant questionnaires, specimen traceability across their organiza- grated and collaborative approach etc. and (2) biomolecular information, tion. As part of our work with the University to solving the informatics chal- including genomics, proteomics, imaging of Texas Medical Branch on our biomedical lenges of translational research. and other high-throughput molecular and informatics systems, we are discussing Contact us to learn more about cellular research data. how our informatics solutions can enable how our solutions for Translational There are many clinical systems capturing translational research collaboration between Research Informatics can help information, many instruments generating both institutions. your organization. ever increasing data volumes, and many Another example is Pfizer. As part of our statistical tools and algorithms available for global agreement to provide informatics analyzing the data, but the key bottleneck solutions to their research groups, we are is contextual data integration from both the supporting Pfizer’s investment in systems clinical and science domains, fused together biology and biotherapeutics. The combi- and structured to be marshaled into analyti- nation of our platform and collaborative cal packages. methodology ensure we design and deploy www.genologics.com an informatics solution that will enable their 1-866-457-5467 Q: Can you provide some examples of translational research and systems biology organizations deploying translational initiatives. research informatics? To connect the clinical and discovery port their translational research efforts. JD: Over the last couple years, we have research processes, the Fred Hutchinson There are many more groups we are seen a tremendous growth in the number Cancer Research Center pursues transla- working with that are conducting transla- of organizations struggling to integrate their tional research under its Translational and tional research, but they each have their informatics systems to achieve translational Outcomes Research Group. We are provision- own unique challenges, which is why an research outcomes. ing the Center with a study management, informatics solution must be highly flexible We have a number of CTSA customers clinical information integration database and and configurable to seamlessly fit with each such as the University of Texas Health biorepository management solution to sup- client’s situation.

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 39 ]

Educational Opportunities

to Medicine, April 22, 2009, This section provides a variety of educational events in the life sciences Washington, DC industry that will help you with your business and professional needs. To list Keynoter Dr. Leroy Hood will address trans- formative technologies taking medicine from an educational event (print & online), contact Sales_chmg@chimediagroup. its currently reactive state to a personalized, com. To preview a more in-depth listing of educational offerings, please visit predictive, preventive and participatory (P4) www.bio-itworld.com mode. Visit www.bioitcoalition.org/events. annual09.html

CHI Events Medical Center, is the featured speaker for Featured Events the lunch meeting at the Arlington Economic For more information on these conferences Development Office, Arlington, VA. Visit www. and other CHI events planned, please visit LabAutomation 2009, January 24-28, bioitcoalition.org/events.html#Jan_09 www.healthtech.com 2009, Palm Springs, CA The Association for Laboratory Automation Electronic Laboratory Notebooks, January Biomarker Assay Development (ALA) presents the 2009 annual LabAutomation 28-29, 2009, London, United Kingdom January 26-28, 2009 * San Diego, CA event. It is the world’s largest conference and SMi’s Electronic Laboratory Notebooks (ELN) exhibition focusing exclusively on the rapidly ASAP Alliance Summit featuring the most qualified speakers in the February 9-12, 2009 * Fort Lauderdale, FL growing field of laboratory automation. Visit industry, share their experiences on imple- www.labautomation.org/LA09 mentation to help attendees make informed decisions. In 2 days, attendees will learn about Molecular Medicine Immunogenicity for Biotechnology, improving knowledge sharing in the chang- Tri-Conference January 26-27, 2009, Washington, DC ing laboratory landscape, the future of the ELN Shaping Future Medicine A hands-on, instructional seminar covering market, and the integration of social software. Molecular Medicine Tri-Conference how to properly design and execute risk- Visit www.smi-online.co.uk/events based immunogenicity programs. Review February 24-27, 2009 * San Francisco, CA operational considerations and learn about TEPR 2009, February 1-5, 2009, Drug Development Latin America emerging trends, new technology and major Palm Springs, CA March 9-10, 2009 * Miami, FL challenges with cell based assays from distin- TEPR is a premier health IT conference for guished industry experts in an intimate, pro- Electronic Data in Clinical Trials healthcare and IT professionals seeking inno- March 11-12, 2009 * Miami, FL fessional environment. Visit www.worldrg vative and practical solutions to current health .com/immunogenicity IT issues. Topics will include, the role of cell Next Generation Sequencing phones in healthcare, linking physicians to March 17-19, 2009 * San Diego, CA Health 3.0: Next Online Generation, hospitals and the patient centered medi- January 26-28, 2009, Las Vegas, NV microRNA in Human Diseases and cal home, and other critical topics related to Development Health 3.0 will review the latest topics on health IT. Visit www.TEPR.com new media, anytime communication and March 23-25, 2009 * Boston, MA optimizing online tools. Analyze large and You May Not Know What You Don’t Know small health plan specific case studies and About Managing Your Trials – learn from speakers representing LinkedIn and Webinar - Thursday, February 5, 2009, Zappos on how to develop online strategies 1:00pm –2:00pm EST and optimize market research. Visit www. Many institutions cobble together bits and worldrg.com/health3point0 Drug Discovery Chemistry pieces of the workflow solution themselves, April 6-8, 2009 * San Diego, CA for example, using Outlook for scheduling, a Meeting the Challenges of Global Clinical series of Excel workbooks for tracking and so Supply Chain Management - on. Many have developed a “home-grown” Webinar - Tuesday, January 27, 2009, system in an often futile attempt to use the PEGS 11:00am – 12:30pm EST “free” IT resources already deployed within the Pharmaceutical companies are conducting hospital to help bring a semblance of order to PEGS – Essential Protein Engineering more clinical trials than ever before, in order the otherwise chaotic task of managing the Summit to spur growth, maintain profit margins and operational aspects of clinical trials. Participate April 6-10, 2009 * Boston, MA replenish shrinking pipelines. The rising costs, in the complimentary Webinar. Visit http:// complexity, as well as regulatory challenges xtalks.com/xto316biooptronics.ashx Advanced Biofuels Summit caused by the increasingly global nature of April 20-21, 2009 * Washington, DC these clinical trials presents significant chal- The World Health Care Congress 2nd lenges for pharmaceutical companies seeking Alliance Management Congress Annual Leadership Summit on Consumer April 22-23, 2009 * Boston, MA to effectively manage the clinical supply chain. Connectivity & Web Empowerment, This Webinar session will cover 10 important February 23-24, 2009, Carlsbad, CA steps to meet these challenges. Visit http:// This summit offers compelling strategies for xtalks.com/xto320catalent.ashx today’s providers and insurers to revolutionize healthcare through the integration and adop- The BIO IT Coalition presents “Systems tion of today’s emerging and, at times, disrup- Bio-IT World Conference & Expo Medicine: A Revolutionary Approach tive technologies. Visit www.worldcongress. April 27-29, 2009 * Boston, MA to the Practice, Business and Future of com/connectivity Medicine”, January 28, 2009, Biomarker World Congress May 27-29, 2009 * Philadelphia, PA Arlington, VA BIO-IT Coalition’s 8th Annual Conference: Dr. Howard Federoff, Executive Vice President Will Biotechnology/IT Save the World? World Pharmaceutical Congress and Executive Dean, Georgetown University The Revolutionary P4 Medicine Approach June 9-11, 2009 * Philadelphia, PA

[40] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

Barnett Educational Services

Visit www.barnettinternational.com for Drug Development & FDA Regulations Examining the Impact of the eCTD on the detailed information on Barnett live seminars, February 26-27, 2009 * Philadelphia, PA Regulatory Submissions Process interactive web seminars, on-site training Conducting Clinical Trials Under ICH GCP February 17, 2009 programs, and customized eLearning develop- March 16-17, 2009 * San Diego, CA Approaches to Address Challenges in ment services. Patient Recruitment & Retention: Vendor Management February 26, 2009 Live Barnett Seminars Successful Planning & Management March 16-17, 2009 * San Diego, CA The Pharmaceutical & Medical Device Medical Device Approval Process Industries Today: Comparing & February 9-10, 2009 * San Diego, CA Good Clinical Practices for Pediatric Trials March 18-19, 2009 * San Diego, CA Contrasting Auditing Techniques for Clinical Research March 3, 2009 Professionals Update on Medical Device Postmarketing February 16-17, 2009 * San Diego, CA Interactive Web Seminars Vigilance Reporting Working with CROs Subject Recruitment for Clinical March 9, 2009 February 16-17, 2009 * Philadelphia, PA Research Sites Critical Decision Points in Design & February 5, 2009 Mastering Cost Management for Global Conduct of Patient Registries Clinical Trials Recent Trends in Noncompliance March 19, 2009 February 18-19, 2009 * Philadelphia, PA February 9, 2009 Operational Modeling & Simulation in Adverse Events: Managing & Reporting Advanced HIPAA: Myths & Current Best Clinical Trials for Pharmaceuticals Practices in Clinical Research April 7, 2009 February 23-24, 2009 * San Diego, CA February 12, 2009

White Papers, Webcasts, and Podcasts

Keep abreast of current industry trends and Software Helps Doping Control Lab Podcasts developments. Browse our extensive list Streamline Results Management Rising Clinical Trial Delays and Costs – of free Life Science white papers, webcasts, Sponsored by Waters Addressing the Cause, Not the Symptoms and podcasts on www.bio-itworld.com. The Karolinska University Sponsored by Medidata Solutions Interested in learning about developing a Hospital’s Doping Control Protocol complexity is taking a toll on clini- multi-media solution to generate leads? Lab tests thousands of cal study speed and efficiency: increasingly Please email Sales_chmg@chimedia samples annually for stimu- complicated and ambitious protocols are not group.com lants, diuretics, and other only burdening sites and study volunteers, but masking agents. Increased also prolonging trials and increasing expenses. regulatory pressure and new White Papers View the podcast to learn how sponsors are techniques increased the responding to this critical issue. 10 Secrets to Recession-Proof number of samples ana- Your Business lyzed creating data management challenges. Next Generation Sequencing Sponsored by Coupa Read about their findings in this case study. Leaders generally face hard Sponsored by GenoLogics choices on how to man- Learn how labs can maximize the value they age a company during receive from their next generation sequencing an economic downturn. Webcasts instruments, such as: This white paper covers 10 Storage for Science – Methods for • What are the data management chal- strategies smart companies Managing Large and Rapidly Growing lenges for labs conducting next generation deploy to recession-proof Data Stores in Life Science Research sequencing their business. Environments • What do genomics labs need to look for Sponsored by Isilon when evaluating lab and data management Webcast covers: Guidance systems High Performance Computing (HPC) in on the many storage Life Science & Education requirements common to Manage Your Biorepositories Sponsored by SGI and Intel Life Science research; The Sponsored by GenoLogics A varied collection of Bio-IT evolution of modern data Learn about solutions for managing your bio- World articles and insights storage architectures; Major specimens and biorepositories such as: are assembled in this data storage architectures • What is the single largest challenge fac- BriefingON that examine key currently in use; and Isilon’s IQ clustered storing organizations in terms of accelerating trends in the HPC infrastruc- age product as a strong and flexible solution to research ture and how researchers those needs. • How is data collection being managed across are putting their best com- disparate organizations and sites putational resources to use. • How will data be managed in terms of com- plying with regulations (e.g., HIPAA)

www.bio-itworld.com JANUARY | FEBRUARY 2009 BIO•IT WORLD [ 41 ]

The Russell Transcript

his group has helped move good projects forward, not just kill bad ones. That gets management’s attention and is no doubt Super Models at a driver of Stanski’s ongoing success. Stanski is scheduled to speak at the 2009 Bio•IT World Conference & Expo in Boston this spring (April 27-29).

IBM Summit Reality Check The conference also contained doses of reality— SAR tools, for example, received decidedly mixed grades from one pre- senter who assessed a fair number of them. It should be said JOHN RUSSELL that many showed accuracy when matched against a specific class of problems. Perhaps the most worrisome note was struck o a considerable extent, the acceptance of modeling during a workshop on modeling in clinical development. At- and simulation (M&S) approaches (beyond tradi- tendees complained that getting executive attention and buy-in tional PD/PK) inside biopharma remains stuck at a was difficult. Top execs are generally not mathematicians and low level and something is needed to get it unstuck— their eyes tend to glaze over when presented too many curves. especially among senior management. Another issue raised was that M&S often delivers bad news. At IBM’s recent, invitation-only Modeling & Killing projects is extremely valuable and saves money but you T Simulation Summit at the company’s Executive Conference don’t get credit for disasters averted. Center in Palisades, N.Y., roughly 85 delegates from around the Measuring modeling and simulation’s contribution to a world gathered for two days of presentations and workshops to project’s success is also a challenge. It’s often not easy to dem- address this question. Most of the presentations demonstrated onstrate that modeling was decisive versus incremental to a solid technical progress and touched on virtually every aspect project’s overall success. Even when modeling is successful, of modeling and every phase of drug discovery and develop- other researchers on the project may feel they would have come ment. IBM, of course, provides both high performance comput- to the correct answer soon enough without modeling. ing technology and has its own ongoing computational biology Yet for many kinds of problems, the M&S kinks have been research efforts. worked out. It’s working at Novartis. It’s being employed on Some projects discussed aren’t ready for prime time but critical projects at Vertex. Biosimulation specialist Entelos soon will be. In one session on heart models, says it has been delivering concrete results distinguished cardiologist and researcher Ar- to clients for years, and while many of those thur Moss from the University of Rochester Showcasing results were project-killers, some were not, reviewed the field to date. Researcher Natalia and co-founder Alex Bangs says there are Trayanova from the Institute for Computational successes can help one or two compounds moving through the Medicine at Johns Hopkins presented visually pipeline whose advance was substantially stunning models exploring cardiac electrome- not only speed helped by M&S. Merrimack Pharmaceuticals chanical functioning. Alejandro Frangi of Fabra M&S adoption, has a compound in phase three and has hung University reviewed EU Heart Project progress. its hat on a discovery-development approach (Conference organizers asked me not to describe but stimulate that relies heavily on mechanistic modeling. presentations in great detail at this time). The best way to overcome barriers to M&S In the Q&A following presentations, Moss creative thinking. acceptance is to showcase successes. While commented on one NIH-funded effort, “This modeling process concern over competitive issues has long kept biopharma tight- may allow early identification medications that prolong the QT lipped about internal projects, the industry as a whole is under interval and mimic the long QT syndrome… to my knowledge such pressure that there is now more openness to different this is the first validation that’s been done in this.” There are of kinds of collaboration and data sharing. It should be possible to course many companies looking at modeling QT prolongation define the elements of a repository of successful modeling and to speed identification of this too-frequent drug killer. Clearly simulations projects in such a way that real competitive issues this is leading edge work whose payoff is still a bit in the future. are not compromised. Conversely, Donald Stanski, global head of modeling and Drug finding and making, it seems to me, is still a mostly simulation at Novartis, discussed the impact his department discovery-informed process rather than engineering-driven; is having today. Stanski’s modeling group is growing, he said, there are many reasonable questions about how fast this pro- and its modeling activities encompass a wide variety of projects cess can evolve. Showcasing successes can help not only speed (preclinical and clinical). Stanski made passing reference to M&S adoption, but also stimulate wider creative thinking having saved $60 million in one area and added pointedly that around improvements to M&S tools.

[42] BIO•IT WORLD JANUARY | FEBRUARY 2009 www.bio-itworld.com

CALL FOR ENTRIES The Bio-IT World Best Practices Awards 2009 “[The Best Practices Awards] is Bio-IT World is accepting entries for its prestigious an indication of how we are all 2009 Best Practices Awards program. Established in working together in this era and 2003, the program recognizes teams for their novel that is a wonderful thing.” and innovative uses of technology, business strategies, –Dr. Francis Collins DQGVROXWLRQVEHQHÀWWLQJWKHELRVFLHQFHVYDOXHFKDLQ Natl Hum Gen Res Inst, NIH from basic research to clinical trials. Direct entries are encouraged as well as nominations from users “Winning the award increased and vendors. our visibility to the external All entries are reviewed and the winners selected by community, especially with regard a distinguished panel of experts from industry and to the rigor with which we do our academia. Awards are presented at a gala dinner held science. That, in turn, enhanced on April 28, 2009, in conjunction with the 2009 Bio-IT our ability to get funding and hire World Conference & Expo, World Trade Center in Boston. people to continue on our growth trajectory.” Best Practices 2009 Entry Categories –Dietrich Stephan, Ph.D. Basic Research & Biological Research Co-founder & CSO, Drug Discovery & Development Navigenics; former Deputy Clinical Trials & Research Director, Translational Genomics Translational Medicine Research Institute Personalized Medicine IT Informatics Knowledge Management Health-IT Manufacturing & Processing

EXTENDED DEADLINE — FEBRUARY 15, 2009 Submit your entry today! Visit www.bio-itworld.com/bestpractices for entry details and submission forms.

Symyx Notebook The Freedom to Experiment.

To learn more, visit www.symyx.com/notebook6