This year’s International Conference on Intelligent Systems for Molecular Biology was the largest gathering of computational biologists ever, according to ISMB organizers. Nearly 2,200 delegates made their way to Glasgow, Scotland, for the meeting, which was jointly held with the third European Conference on Computational Biology, July 31-Aug. 4.
The record-breaking attendance mirrored the expanding scope of the conference: For the first time, papers were presented in parallel tracks in order to cover a broader range of topics. In addition to the usual lineup of genome and transcriptome analysis, phylogeny, protein analysis, and pathways, there were sessions devoted to databases, ontologies and text mining, and biomedical applications.
Janet Thornton, director of the European Bioinformatics Institute and program chair for the conference, said 200 reviewers selected the 67 papers from 492 submissions. In addition, an unprecedented number of posters — nearly 900 — covered a similar spectrum of topics, and the special interest group meetings that preceded the main conference also ballooned this year, to 10 meetings spanning two days.
Enthusiasm was high, particularly for some of the newer topics: The ontology and text mining session drew such a crowd on the first day that organizers had to scramble to provide two overflow rooms, and some of the software demonstrations attracted upwards of 200 people — more than the total number of delegates at the first ISMB, held in Bethesda, Md., in 1993.
Larry Hunter of the University of Colorado School of Medicine, who was at that inaugural ISMB conference, said he was pleased to see so many new faces at this year’s meeting. In addition, he told BioInform that it was gratifying to see topics like graph theory and ontologies working their way into the common parlance of bioinformatics. ISMB originally grew out of the artificial intelligence community, he noted, but that field’s influence had “faded away” in the intervening years. A better understanding of ontologies and other AI concepts, he said, “will be crucial for this field.”
Phil Bourne of the University of California, San Diego, also remarked that core computer science concepts are becoming more integrated into bioinformatics. In previous years, he told BioInform, the split between biologists and computer scientists was clearly evident at ISMB, but the two communities appear to have found some common ground.
“It’s a reflection of the beginning of the maturity of the field,” he said. In addition, he noted, there was a particular “vibrancy” at this year’s conference because many of these key computer science methods are helping scientists “march up the steps of biological complexity.”
As one attendee summed it up, “They put the ‘IS’ back into ISMB this year.”
The recent emergence of systems biology has clearly driven computational biologists to seek out new methods for representing and analyzing biological data — a theme that provided a logical framework for the ISMB 2004 program.
Lee Hood of the Institute for Systems Biology kicked the meeting off with an overview of his work and a look at some specific computational challenges in the field, while David Lipman closed the conference with the ISCB Senior Scientist Accomplishment Award Lecture, “Message and Meaning in Sequence Comparison: Is Systems Biology Possible?”
Lipman warned against unrealistic expectations when analyzing biological systems above the level of single genes. Evolution often compensates for genetic substitution or loss, he said, so that even the smallest genetic change can have far-reaching and unforeseen phenotypic results that are still too complex for computational approaches to understand. There are some “exciting things coming out of systems biology,” he said, “but to approach these systems without a deep understanding of evolution will be difficult.”
Other keynotes also addressed approaches for grappling with many levels of biological complexity: Oxford’s Dennis Noble discussed how he is using computational systems biology to integrate molecular-level data into organ-scale simulations of the heart; Eric Green of the National Human Genome Research Institute spoke about recent advances in algorithms for comparative genomics that are beginning to make some sense of non-coding functional regions of the genome; and Uri Alon of the Weizmann Institute and winner of this year’s Overton Prize discussed methods he has developed to simplify the representation of biological networks using “network motifs” — three-node and four-node recurring patterns that can reduce the complexity of large, complex networks to help create a simplified “blueprint” for biological mechanisms. “Knowing that blueprint will give us unprecedented medical ability,” he said.
… and Networking
Many attendees noted, however, that the value of ISMB isn’t so much in the scientific quality of the papers, but in the social interactions that the meeting enables. Indeed, the number of peripheral activities at this year’s meeting kept some from attending many talks at all.
In addition to the 10 SIG meetings prior to the main meeting, a large number of “birds of a feather” sessions took place during the conference to address a range of issues and topics in the field, such as standards, new database initiatives, semantic web technology, funding, and the like. Vendors and academic groups also kept attendees jumping between two separate software demo tracks during the main session.
Poster receptions were very well attended, with delegates navigating a sea of nearly 1,000 presentations, cocktails and hors d’oeuvres in hand. The socializing continued long after the talks ended each day, and more than a few brave souls even made their way to the dance floor following the gala dinner to kick up their heels in some traditional Scottish Ceilidh dancing.
The high attendance at this year’s conference raised more than a few questions about whether ISMB has gotten too big, but the turnout was largely due to the fact that it was jointly held with ECCB. Next year, organizers noted, the meetings will be held separately again:
ISMB 2005 will be held in Detroit, Mich., June 25-29, while ECCB 2005 will be in Madrid, Sept. 27-Oct. 31.
ISMB 2006 is scheduled for Aug. 6-10 in Fortaleza, Brazil.