Skip to main content
Premium Trial:

Request an Annual Quote

Pathway Informatics Gains Ground at ISMB 05: Resources Grow as Support for BioPax Widens


DETROIT — A surge in new databases and analysis tools bumped pathway informatics from niche status to mainstream fare at this year's Intelligent Systems for Molecular Biology conference.

More than 1,650 computational biologists attended ISMB 2005, held here last week, making the conference the largest "free-standing" ISMB ever, according to Michael Gribskov, president of the International Society for Computational Biology, the sponsoring organization. Last year's meeting, which was jointly held with the ISCB's annual European Conference on Computational Biology, attracted around 2,200 delegates.

Conference activities spanned four floors of the 70-story Marriott Renaissance Center, as delegates navigated the hotel's labyrinthine layout to check out three parallel paper sessions, four parallel software demonstration tracks, tutorials, workshops, and more than 700 posters.

But talks, demos, and posters associated with pathway databases and analysis tools tended to attract the most attention this year — a trend that didn't go unrecognized among conference exhibitors. PubGene, Ariadne Genomics, Lion Bioscience, Biobase, and Gene Logic were all pitching products related to biochemical pathways (see ISMB Conference Briefs, this issue, for further details on vendor offerings).

Gary Bader, a postdoc in the Computational Biology Center at Memorial Sloan-Kettering Cancer Center, said that he and his colleagues have identified more than 180 pathway databases and analysis tools in the public domain (a full list is available at

The question of how — or whether — pathway data can be shared is entirely up to the database developers, who need only "make their data available in a sharable format." The bigger question today is, "'How do we develop pathway databases?'"

Bader, who has helped develop the BioPax exchange format for pathway data, spoke about the standards effort at the BioPathways special interest group meeting that preceded the main conference. He said that support for BioPax is growing within the pathways community. For example, BioCyc data is now available in BioPax level 1, which represents metabolic pathway information. KEGG is preparing to release its data in BioPax level 1 "soon," Bader said, and WIT/Puma2, Reactome, and aMaze are expected to be available in BioPax some time this summer. A number of pathway analysis tools, including cPath, Cytoscape, GenMapp, Patika, Qpaca, and VisAnt, also support the standard, Bader said.

BioPax level 2, which will address protein-protein interactions and post-translational modifications, will be available before the end of the year, Bader said, and the BioPax consortium has targeted level 3 — which will support molecular states and gene regulation — for early 2006. Level 4 — which will address genetic interactions and "generic entities" — is slated for late 2006.

The standard has certainly made a great deal of progress over the last year. In contrast to last year's BioPathways SIG, in which there was a good deal of spirited debate about the best means of sharing pathway data [BioInform 08-09-04], participants in a panel discussion on the same topic this year seemed to have arrived at a consensus that BioPax will suit the community's needs.

Peter Karp of SRI, who maintains the EcoCyc and BioCyc databases, noted that the question of how — or whether — pathway data can be shared is entirely up to the database developers, who need only "make their data available in a sharable format." The bigger question today, Karp said, is "'How do we develop pathway databases?'"

Indeed, the conversation quickly turned to the technical and financial challenges of developing and maintaining these resources, which require a large amount of manual curation — a time-consuming and labor-intensive process that, at present, cannot be replaced with computational methods.

Eric Andrade of the Blueprint Initiative, which has been struggling to secure follow-on funding from the Canadian government for its BIND resource [BioInform 05-09-05], said that it's important for pathway resources to "find a jurisdiction that understands that this is a library and funds it as such." Blueprint, which also has funding from the government of Singapore, is still in discussions with Canadian funding agencies regarding continued support, Andrade said.

In another sign that the pathways field is developing rapidly — and getting a bit more competitive — a number of speakers took aim at one of the pioneering efforts in the area, the KEGG database at Kyoto University. During the BioPathways SIG, SRI's Karp provoked a good deal of discussion by comparing KEGG with BioCyc to show that KEGG pathways are "larger" than those in BioCyc, indicating that the resource incorrectly combines biologically separate processes into single pathways.

In a poster abstract that was elevated to an oral presentation, Tsai-Tien Tseng of the University of Illinois at Urbana-Champaign noted that KEGG was "problematic." The resource lacked "a couple dozen" key enzymes, Tseng said, even though the upstream and downstream enzymes within those pathways were present. As a result, Tseng and his colleagues had to re-annotate the bacterial genomes they were studying to reconstruct the pathways.

Ewan Birney of the European Bioinformatics Institute also pointed out some drawbacks with KEGG during his Overton Prize keynote address on the final day of the conference. Birney said that the Reactome database of human pathways that EBI is creating in collaboration with Cold Spring Harbor Laboratory was born, in part, from "frustration" with KEGG. Reactions in KEGG are linked only by EC numbers, he said, so it's difficult to tell if two proteins are acting separately or as a complex. "We wanted to collect data in a more careful, correct way," he said.

The challenge, he noted, is that "there's no ABI sequencer for pathways, so we're plundering people's brains."

Reactome currently contains pathway information for "just shy" of 1,000 proteins, Birney said, and is expected to scale "to capture the majority of pathways that are known in molecular biology."

— Bernadette Toner ([email protected])

Filed under

The Scan

Transcriptomic, Epigenetic Study Appears to Explain Anti-Viral Effects of TB Vaccine

Researchers report in Science Advances on an interferon signature and long-term shifts in monocyte cell DNA methylation in Bacille Calmette-Guérin-vaccinated infant samples.

DNA Storage Method Taps Into Gene Editing Technology

With a dual-plasmid system informed by gene editing, researchers re-wrote DNA sequences in E. coli to store Charles Dickens prose over hundreds of generations, as they recount in Science Advances.

Researchers Model Microbiome Dynamics in Effort to Understand Chronic Human Conditions

Investigators demonstrate in PLOS Computational Biology a computational method for following microbiome dynamics in the absence of longitudinally collected samples.

New Study Highlights Role of Genetics in ADHD

Researchers report in Nature Genetics on differences in genetic architecture between ADHD affecting children versus ADHD that persists into adulthood or is diagnosed in adults.