SAN JOSE, Calif.--The major outcome of a meeting of the Object Management Group's Life Sciences Research Domain Task Force, held here late last month, was the final submission of a proposed CORBA standard for biological sequence analysis software. Concept Five Technologies, the European Bioinformatics Institute, Genome Informatics, Millennium Pharmaceuticals, Neomorphic Software, and NetGenics presented the standard jointly. After a vote, which is scheduled to take place at the next meeting in Cambridge, Mass., November 15-19, the OMG will issue an official standard. It will be the first produced for life sciences research.
The task force is also making progress on other standards. Douglas Greer, chair of a working group on macromolecular structures software, issued a request for proposals for a standard in that area. Greer said the specification would be a first-step to define what structure information can be retrieved with a CORBA interface. "As more companies and researchers become involved with CORBA, it's important to have an interface, in particular to the Protein Data Bank, that keeps track of protein structures," he explained.
Greer works for the Research Collaboratory for Structural Bioinformatics at the San Diego Supercomputing Center whose charter includes developing a CORBA interface for the research community. "This would be an important interface for pharmaceutical companies and other commercial entities," Greer said. "That's why I'm involved in the OMG--we're trying to get a proposal together that would meet both needs."
But pharmaceutical companies haven't been as forthcoming about their software needs as many members of the task force would like. Work on some standards has idled at the information-gathering stage for want of further input from software users, in particular pharmaceutical companies.
Alex Bangs of Entelos, said the working group on bibliographic services software, which he chairs, elected to extend its deadlines in hopes of getting feedback from a broader population. "People like NetGenics or Concept Five are producing software but not the actual data. We'd like to hear from people on the data side, or convince the US National Library of Medicine to participate," Bangs said.
Similarly, a request for input on a gene expression standard that was issued nearly a year ago has received three responses, all from developers of gene expression products. Douglas McArthur, chair of the gene expression working group, said, "We've extended the deadline again because we're really hoping to get some more responses from pharmaceutical companies, not just vendors."
McArthur blamed several factors for poor participation in gene expression standard-setting, including "a major player for whom it is not in the best interest to have standards." Users of Affymetrix's GeneChip, he said, are locked into using its packaged software, and many are not yet at the point where they need other systems. "Vendors anticipate the need because we've seen this pattern with high-throughput screening and other bioassay-type automated experiments."
But, he acknowledged, pharmaceutical companies might also be reluctant to offer input for fear of revealing trade secrets. McArthur said his employer, Rosetta Inpharmatics, is involved in the standards-setting because it foresees users of its own gene-chip technology, which uses an ink jet printer head to print base pairs, wanting to integrate data collected from different kinds of chips. "That's where the interoperability standards really come into play," he noted.
--Adrienne Burke