While the European Bioinformatics Institute, NetGenics, and Rosetta Inpharmatics have agreed to work together through the Object Management Group on a joint submission proposal for a gene expression standard by mid-June, tensions may run deeper than appearances suggest.
Ugis Sarkan, an EBI database administrator involved in the group’s Microarray Markup Language, responded to Rosetta’s recent public relations campaign trumpeting industry support of its Gene Expression Markup Language standard and the Nature Publishing Group’s decision to sign on as a member of the GEML consortium.
He pointed out that the organizations that have agreed to use GEML so far, including Spotfire, Agilent Technologies, Europroteome, and the Harvard University Center for Genomic Research, "use mostly Rosetta-type platforms."
He said that MAML, by contrast, is platform independent. In addition, he claimed, "MAML has a wider support in academia and a more generic approach."
According to Sarkan, the MAML community recently sent a letter to Nature "pointing out that currently MAML has a wider support, especially in academia but also in industry."
Douglas Bassett, senior director of computational biology at Rosetta, declined to comment on the company’s plans to integrate its proposal.
The next forum for debate on the standards will be at the Pacific Symposium on Biocomputing, where the BioPathways Consortium will take a first look at adopting GEML, said Eric Neumann, 3rd Millennium’s vice president of life science informatics and a co-founder of the consortium.
"We are going to look into what its value is say over EBI’s MAML. But that you can’t answer in one meeting. It’s not going to be that critical for us to answer that right away," said Neumann, who along with Peter Karp and Pedro Romero of SRI International’s artificial intelligence center, is directing the session on genomic, pathway, and interaction data on January 6.
Neumann said that GEML is a little more streamlined and less complex than MAML, which might make GEML easier to use.
Other gene expression data format standards are also being developed by the National Center for Biotechnology Information, which has launched a database called Gene Expression Omnibus.
It is too soon to say what is the best standard, said Neumann, adding that the evolving standards are very important for better understanding biological pathways.
"Pathways are, to a large extent, one of the major recipients of micro- array data. We want to be able to handle that and all the inferencing that’s needed with it," said Neumann.
Overall, the session being conducted by Neumann will give bioinformaticists a chance to discuss developments and requirements for computational tools for modeling and managing metabolic pathways, signal-transduction pathways, genetic regulatory circuits, and protein-protein interactions.
The morning section will feature presentations by authors of papers and discussions, including a panel talk. In the afternoon, members of the BioPathways Consortium will present on preliminary proposals for potential standards, he said. The consortium’s six-month agenda will also be mapped out.
Other presentations will come from Physiome, which will discuss its CellML markup language, which, according to the company, enables people to model various cellular processes. And Transpath, an organization whose members have been investigating bacterial gene regulation, will give a talk on its research.
After the scientific talks are done, the consortium will host a party sponsored by BioCarta, Roche Diagnostics, 3rd Millennium, and Sun Microsystems, who are all sponsors of the consortium.
The next BioPathways meeting will take place at the Intelligent Systems for Molecular Biology in Copenhagen on July 21-25. The group expects to have some recommended standards to present at that gathering.
—Aaron J. Sender and Matthew Dougherty