While delivering his paper on normalized sequence alignment at Recomb 2001, Pavel Pevzner noted that Temple Smith and Michael Waterman were not the first to address the problem of sequence alignment, but they were the first to “get the question right” in order to resolve it with their famous algorithm.
Computational biologists working on the next generation of algorithms will have no shortage of questions to ask, it seems, as Recomb’s invited speakers encouraged attendees to apply computational approaches to a growing number of emerging biological challenges.
Over 600 people from 27 countries attended the Fifth Annual International Conference on Computational Molecular Biology, held in Montreal, Canada, April 22-25, nearly doubling the attendance of last year’s meeting in Tokyo. Participation was truly interdisciplinary — computational biologists and molecular biologists were joined by a healthy mix of computer scientists, mathematicians, statisticians, and physicists. Of the attendees, 28 percent were from industry, a marked increase from previous years, while 57 percent were academics, and 14 percent were from government agencies.
While the meeting’s 35 contributed papers focused on mathematical and computational methodology, the nine keynote talks provided a higher-level view, placing the algorithmic work within the context of wet lab biology, systems biology, and the drug discovery process.
Philip Sharp of the Massachusetts Institute of Technology explained the recently discovered process of RNA interference, which may silence repetitive sequences in the genome. Mark Ptashne of Memorial Sloan Kettering Cancer Center discussed his current research on gene regulation, which indicates that intermediate activators play a role in “recruiting” biologically important enzymes. This interactivity would then determine binding activity, as opposed to specific active sites. Franz Lang of the University of Montreal spoke about mitochondrial genomics, and directly challenged attendees to develop better computational means to identify structural RNA genes.
In his talk on the current limitations of functional genomic technologies, Harvard’s George Church called for a “Biosystems Project” that would compare with the Human Genome Project. He stressed the need for sharable, integrated models of cellular systems. Roger Brent of the Molecular Sciences Institute expressed similar ideas in his discussion of information processing by cells. Brent proposed developing “Biological Information Systems” modeled after Geographic Information Systems that could be used to predict biological activity. Matthias Wilm of the European Molecular Biology Laboratory suggested that a protein complex characterization technique developed at EMBL could be used for protein-protein interaction mapping, which could then result in a virtual cell to simulate biological processes.
Yvonne Martin of Abbott Laboratories and Klaus Lindpainter of Hoffmann-La Roche addressed the impact of computational biology on the pharmaceutical industry. Martin discussed how computational chemistry could be used to predict the behavior of molecules before they are synthesized, while Lindpainter delivered an overview of the impact genetics and genomics will have on drug discovery and development.
Mark Adams of Celera Genomics discussed Celera’s sequencing and subsequent analysis of the human genome.
Contributed papers addressed protein structure, molecular interactions, expression patterns, sequencing by hybridization, sequencing and sequence analysis, phylogeny and gene duplications, and proteomics and RNA structure. There were also 130 scientific posters.
Recomb 2001 organizers were pleased with the impressive turnout and the quality of the papers presented. Thomas Lengauer, program chair, said that most of the papers “addressed immediately application-relevant problems and validated their methods on real data, so as to make direct contributions to biology.” Conference chair, David Sankoff of the University of Montreal, noted that the right mix of new mathematical work and biological applications is considered key to the success of the conference.
Indeed, attendees expressed enthusiasm for the interdisciplinary nature of the talks and for the balance between computer science and biology. Because so many recent recruits to computational biology lack training in biology, the invited speakers helped serve as a reminder that “these aren’t just elegant mathematical problems to solve,” as one participant remarked.
Recomb 2001 was hosted by the Centre de Recherches Mathématiques and was sponsored by the Association for Computing Machinery with additional support from several industry, government, and trade organizations.
Recomb 2002 will be hosted by Celera Genomics and will be held in the Washington area.