Comparative and Functional Genomics Comp Funct Genom 2003; 4: 654–659. Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/cfg.336 Meeting Review 11th Intelligent Systems for Molecular Biology 2003 (ISMB 2003) Brisbane Convention Centre, Brisbane, Australia, 29 June–3 July 2003 Catherine A. Abbott* School of Biological Sciences, Flinders University, Adelaide, SA, Australia *Correspondence to: Abstract Catherine A. Abbott, School of Biological Sciences, Flinders This report profiles the keynote talks given at ISMB03 in Brisbane, Australia by Ron University, GPO BOX 2100, Shamir, David Haussler, John Mattick, Yoshihide Hayashizaki, Sydney Brenner, the Adelaide, SA, Australia. Overton Prize winner, Jim Kent, and the ISCB Senior Accomplishment Awardee, E-mail: David Sankov. Copyright 2003 John Wiley & Sons, Ltd. cathy.abbott@flinders.edu.au Received: 22 September 2003 Revised: 25 September 2003 Accepted: 29 September 2003 Introduction Evolutionary Analysis’ and ‘Molecular Modelling: building a 3D protein structure from its sequence’, This year for the first time the annual meeting of and on more cutting-edge topics such as ‘Bioethics the International Society of Computational Biol- for Bioinformaticists’, ‘Artificial Intelligence and ogy (ISCB) was held in the Southern hemisphere. Machine Learning Techniques for Bioinformatics’ The 11th International Conference on Intelligent and ‘Data Warehousing in Molecular Biology’. Systems in Molecular Biology (ISMB2003) came The official meeting began on Monday with a ‘down under’ and was held at the Brisbane Conven- warm welcome by the co-chairs of the meeting, tion Centre, Brisbane, Australia, 29 June–3 July Gene Myers and Mark Ragan, to the 919 registrants 2003. Unfortunately, the promised great Aussie from 42 different countries. This review covers the sun did not come out as much as was expected keynote and prize-winning talks that were delivered but the programming committee delivered a stim- across the 4 days of the meeting. All other talks are ulating program where the ‘bio’ was once again published as papers in the journal Bioinformatics returned to bioinformatics. (http://bioinformatics.oupjournals.org/). Notable As is traditional at ISMB, the meeting was speakers at ISMB 2003 included: Sydney Bren- preceded by Special Interest Group (SIGS) satel- ner (one of three recipients of the 2002 Nobel lite meetings and tutorials. These SIGS were well Prize in Medicine, founder of the Molecular Sci- attended, with the numbers of participants rang- ences Institute and Distinguished Research Profes- ing from 98 for the Bioinformatics Open Source sor at the Salk Institute); David Haussler (Howard Conference, 100 for Biopathways, 51 for Bio- Hughes Medical Institute Investigator and Pro- Ontologies, 28 for WEB03, and 42 for Text Mining fessor of Computer and Information Sciences at (BioLINK). The tutorials were popular, as usual, the University of California, Santa Cruz); Yoshi- with over 500 ISMB registrants taking part in the hide Hayashizaki (Project Director of the Genome 15 tutorials offered. These were held on tradi- Exploration Research Group at the Genomic Sci- tional topics, such as ‘Molecular Phylogenetics and ences Center at RIKEN); John Mattick (Director of Copyright 2003 John Wiley & Sons, Ltd. Meeting Review 655 the Institute for Molecular Bioscience, University One of the grand challenges of human molecular of Queensland); Ron Shamir (Professor of Com- evolution is to reconstruct the evolutionary history puter Science at Tel-Aviv University); and Michael of each base in the human genome. Great sums of Waterman (Professor of Mathematics, Computer money were spent on getting a draft sequence of Science and Biological Science at the University the human genome, so we now need to use this of Southern California). information fruitfully. Genome browsers, such as This year’s ISMB focused more on the use of that at UCSC, will become more than archives of bioinformatics and computational biology to anal- data — they will be used as microscopes that can yse entire biological systems and there was less be used to interpret and discover new things about emphasis on individual standalone problems, such human genes. The overall goal of future browser as microarray analysis, building phylogenetic trees, applications is to link gene position to functional motif- and gene-finding algorithms. Papers and information, so that the browser can be used as keynotes at the conference were divided into seven an engine for discovery — a microscope searching major themes: (1) phylogeny and genome rear- the genome for new and exciting discoveries. These rangements; (2) expression arrays and networks; thoughts were expanded upon later in the meeting (3) predicting clinical outcomes; (4) protein clus- in the Overton Prize address by Jim Kent. tering, alignment and patterns; (5) transcription The second keynote address in this session motifs and modules; (6) structure and hidden was John Mattick’s (University of Queensland; Markov models; and (7) text mining and high- http://www.imb.uq.edu.au/mattick.html), ‘Pro- throughput methods; with an additional session for gramming of the autopoietic development of com- short papers. plex organisms: the hidden layer of non-coding With the sequences of over 1000 genomes now RNA’. His presentation generated enormous excite- completed, including those of human [13] and ment in the audience and attracted an incredible mouse [20], the meeting opened with a thought- amount of interest in the form of questions after- provoking session on phylogeny and genome rear- wards. rangements, which included two keynote speakers. The number of protein-coding genes does not David Haussler (University of California; scale strongly with the complexity of an organism. Santa Cruz; http://www.cse.ucsc.edu/∼haussler) The fly and worm genomes contain 14 000–19 000 gave a fascinating presentation on ‘Identifying protein-coding genes, two to three times more than functional elements in the human genome by yeast at ∼6000, but this is not much less than the tracing the evolutionary history of the bases: a number of mammalian protein-coding genes. More- key challenge for comparative genomics’. He out- over, the genome of the inanimate plant, rice, has lined the principles of using evolution to find more genes than that of humans. The 98% of non- genes and other functional elements [19]. Although coding sequence in the human genome actually 75 million years of evolution separate mice and has a tighter correlation with human complexity. humans, comparing the genomes of the two species Most of the genetic variation between organisms provides a crude way of finding regions of func- lies in these non-coding regions, indeed only 1% of tional significance. This type of approach is noisy coding genes between mouse and human are dif- and, while difficult, it is possible to separate the ferent. Mattick gave a compelling argument that noise from the useful information using a cali- non-coding RNA is the genetic basis of human bration point. At least half of the human genome complexity and variation [14,15]. There are enor- consists of relics of retrotransposons, i.e. half of mous numbers of non-coding RNA genes in the our genome is the rotting carcasses of the selfish mammalian genome, which are only now begin- DNA that has inserted itself into our DNA over the ning to be recognized, and which appear to account years. Comparative genomics will allow us to iden- for between one-half and three-quarters of all tran- tify functional elements and, as more species are scripts. At least 50%, and possibly the majority, sequenced, there will be an increase in our power to of the human genome is transcribed. He fielded detect conserved elements. His group has enhanced the hypothesis that RNAs derived from processed these studies by combining hidden Markov and introns are involved in gene–gene communication phylogenetic models to create new ways of mod- and networking in real time in eukaryotic cells. elling molecular evolution [17]. The talk summarized the accumulating evidence Copyright 2003 John Wiley & Sons, Ltd. Comp Funct Genom 2003; 4: 654–659. 656 Meeting Review that spliced RNA molecules are functional; introns expression data include the new biclustering algo- comprise, on average, 95% of the primary sequence rithm SAMBA (Statistical Algorithmic Method for of protein-coding transcripts in humans, the num- Bicluster Analysis) and PRIMA (PRomoter Inte- bers and size of introns and non-coding RNA cor- gration in Microarray Analysis), a program for relates with developmental complexity, and some finding transcription factors whose binding sites are introns/non-coding RNAs are highly conserved. enriched in a given set of promoters. By utilizing The emphasis on systems biology was contin- human genomic sequences and models for bind- ued in the next session on expression arrays and ing sites of known transcription factors, PRIMA networks. While there were many interesting pre- identifies transcription factors whose binding sites sentations, the one that clearly stood out was are significantly over-represented in a given set that of Eran Segal (Stanford University; www- of promoters. Another JAVA-based tool his group cs.stanford.edu/∼eran) on ‘Discovering molecu- has developed to aid clustering and visualizing of lar pathways from protein interaction and gene gene expression data is EXPANDER (EXPression expression’, which received the award for Best Stu- ANalyzer and DisplayER). This visualization