Standards for life science identifiers, biological pathways, SNPs, and gene expression database interfaces were on the table at the Object Management Group’s recent meeting in Boston, Sept. 8-12 — and a few of them are already inching their way toward adoption, according to Dave Benton, co-chair of the OMG’s Life Science Research domain task force.
Benton told BioInform that the OMG LSR currently has six active requests for proposals in its standardization pipeline: “Life Science Identifiers,” “Biochemical Pathways,” “Gene Expression Query Service,” “Single Nucleotide Polymorphisms,” “Life Science Analysis Engine,” and “Compound Collections.”
At the Boston meeting, Lion Bioscience and France’s Centre National de la Recherche Scientifique (CNRS) submitted initial proposals for the Biochemical Pathways standard. Lion’s proposal (available at http://www.omg.org/cgi-bin/doc?lifesci/03-08-01) is based on the existing CellML and SBML XML-based markup languages for pathway reconstruction, while the CNRS proposal (available at http://www.omg.org/cgi-bin/doc? lifesci/03-08-03), dubbed SB-UML, is based on the OMG’s Unified Modeling Language (UML). Benton described the two proposals as “complementary,” because SB-UML would act as a “meta-model that can be mapped to the existing XML schemas” of CellML and SBML. Lion’s proposal would essentially act as a subset of the CNRS model, he explained.
OMG is not the only organization sponsoring a standards effort for pathway data. BioPAX, a consortium of several biopathway databases and research groups, is also working on a unified data standard for the field. Joanne Luciano, a member of BioPAX who was at the OMG meeting, told BioInform that BioPAX is already working very closely with the SBML and CellML efforts, and that it has also had several discussions with the Lion developers working on the OMG proposal. The CNRS proposal highlighted “a number of biological concepts that we want to make sure are in BioPAX,” she said, so communication between those efforts is also likely. It’s doubtful, however, that BioPAX will play an active role in the OMG process, Luciano said. While the initiative plans to keep its work “in sync” with other efforts, “we have limited resources and want to concentrate on getting the bugs out of BioPAX … If they want to adopt us after the fact, we’re more than happy to be adopted when the time comes,” she said.
Also on the agenda in Boston was an evaluation of the LSID specification that the Interoperable Informatics Infrastructure (I3C) submitted against the OMG’s Life Science Identifier RFP. Benton said that IBM and the European Bioinformatics Institute also submitted a proposal for an LSID “resolver” service that will allow researchers to retrieve biological objects from distributed resources using the LSID syntax. “It’s not clear to me that the I3C has made progress toward coming to a final recommendation on the resolver part,” Benton said, but he added that since this is the first example of a joint standard development between the I3C and OMG LSR, “we’re still finding our way.”
On the SNP standard front, Benton said there was an “encouraging progress report” from the Japan Biological Information Consortium, which is coordinating an effort among the major SNP databases to work out a standard that they plan to submit to the OMG at its February meeting in Anaheim, Calif.
The OMG LSR’s next meeting is scheduled for Nov. 17-21 in London, where the group will hear initial submissions against the Life Science Analysis Engine RFP, which seeks to standardize access to analysis tools in the life sciences (available at http://www.omg.org/cgi-bin/doc?lifesci/2003-01-08). Initial submissions for the Compound Collection RFP will also be accepted at the London meeting, along with revised submissions for the LSID standard.