LEESBURG, Va., Oct. 8 – If there is to be a Human Proteome Project, Francis Collins told a group of proteomics researchers Monday, it will most likely occur in “bits and pieces,” with academic and industry researchers working together to complete specific goal-oriented projects.
As one of the keynote speakers at “Defining the Proteomics Agenda,” a conference organized by the American Chemical Society held at a golf resort here, Collins sought to draw upon his experience leading the public effort to sequence the human genome by encouraging the assembled researchers to forge public/private partnerships, among other suggestions, as an efficient way to accomplish large data-collection initiatives.
In addition, Collins told the 250-odd researchers that it was important to develop the required technology for a large scientific project before scaling up and devoting large resources to complete it, and that encouraging interactions across traditional disciplines would help stimulate the necessary technology development. Predictably, Collins said any form of proteome project should release all data to the public without restrictions.
“For the human genome project, there were very concrete deliverables,” he said. “By such-and-such a time, we will have such-and-such data set. We’re struggling now to come up with a parallel set [for proteomics].”
But public/private partnerships that focus on specific milestones within a certain timeframe, and that generate pre-competitive data, he added, could serve both academic and industry proteomics researchers well. Such projects would require robust oversight, equal financial contributions from both public and private sources, and rigorous quality control, he said.
“If you can come up with a variety of data sets in proteomics that fit these criteria I would expect they would be interesting partnerships,” he said. Collins cited the SNP consortium and mouse genome sequencing effort as good examples of this kind of partnership in genomics.
In the short term, however, Collins said the National Human Genome Research Institute would try to address the scarcity of public proteomics databases by issuing a request for proposals within the next month to fund large protein database projects. Collins told GenomeWeb NHGRI was looking to fund “one large database” already in existence, rather than a project designed to create several new protein databases.
When asked what new technology would be most important to develop for proteomics, Collins said he could not limit himself to just one, and cited techniques for studying protein interactions, protein microarrays, and separation methods as equally deserving of attention. “Clearly we need a collection of front ends for the mass spec to do [protein separation] at high throughput,” he said.
Although finding solutions to these technical issues would not be easy, Collins encouraged the attendees to steel themselves for the task. “The human proteome is still a challenge but this is a challenge we should not shy away from.”