Questions of biological and clinical relevance predominated at the Human Proteome Organization's 10th annual meeting this week in Geneva, as much of the discussion focused on efforts to move the field of proteomics in more applied directions.
Most significantly, the organization voted to add a new arm to the Human Proteome Project that it launched last year. Dubbed the Biology/Disease-Driven Human Proteome Project, or B/D-HPP, this new arm aims to organize characterization of the proteome based on proteins' roles in disease and biological functions. In addition, it will complement the chromosome-centric approach the organization has been pursuing for the last year.
The B/D-HPP was proposed to the organization by University of Michigan professor Gil Omenn and Swiss Federal Institute of Technology Zurich researcher Ruedi Aebersold. It stems from the notion, Aebersold said in a presentation to the congress, that, while currently, "most biological research is focused on [individual] biological molecules … that is changing very quickly, and the focus is clearly shifting to networks, especially in the field of proteomics where proteins clearly act in coordination with other proteins to convert the phenotype."
Thus far, Aebersold noted, the HPP's work has largely confined itself to the proteins described in the UniProt database, many of which have been derived from genome-sequencing studies. Now, he said, the organization needs to move to develop a better understanding of phenomena like "sequence variants, splice forms, post-translational modifications, functional annotations, [protein-protein] interactions and their significance, and where proteins are expressed and under what conditions."
Given the challenge of investigating these questions on a proteome-wide scale, it's necessary and desirable, Aebersold said, "to slice up the proteome into specific groups of proteins of interest so that issues like these can be studied in detail." The B/D-HPP, he added, provides "a way to group proteins into interesting groups based on disease mechanisms, which can then be studied."
This move toward a biology- and disease-driven approach fits with the progression of the chromosome-centric human proteome initiative, or C-HPP, that was launched at last year's meeting (PM 2/18/2011).
Calling for participating countries to take one of the 23 human chromosomes and characterize one representative protein for each gene located on the chromosome, the C-HPP is not explicitly disease-focused.
However, as Young-Ki Paik, director of the Yonsei Proteome Research Center in Seoul and chair of the C-HPP consortium told ProteoMonitor in February, disease concerns have driven many participants' chromosome selections. For instance, Japan chose to focus on chromosome 3 in part because of its association with lung cancer, while Sweden selected chromosome 19 due to its link to prostate cancer, a research interest of several of the team's scientists.
The C-HPP effort has lined up teams for 14 of the 23 chromosomes, with seven new groups added since last year's meeting. During presentations this week detailing the teams' progress, another reason for the shift toward questions of disease and biological processes became apparent: the need for funding.
Macquarie University professor Mark Baker, leader of Australia's investigation into chromosome 7, noted that one of the major issues his team had faced was that "it was difficult to get a basic science project primarily just around chromosome 7 funded in any way, shape, or form" without some link to its impact on human health.
Charles Pineau, director of the Biogenouest Proteomics Core Facility in Rennes, France, and one of the leaders of France's study of chromosome 14, likewise noted that "manpower and funding" were primary concerns. He added that the team aimed to include a disease-driven component to its work by reaching out to investigators studying ailments linked to the chromosome.
"HUPO is for setting policies and providing a framework, but it doesn't have the resources to come up with the money," Aebersold said, citing funding concerns as a key rationale for the B/D-HPP proposal. "As scientists who want to execute some of these projects, we need to be able to go to our funding agencies and say, 'This looks like a great idea and an important project. Can you fund it?'"
"We think that if we can present the Human Proteome Project in a disease context, it will be easier to resonate with funding agencies and maybe also the private sector to get these projects viewed positively and to get the resources to actually do the work," he added.
Lund University professor and AstraZeneca principal scientist Gyorgy Marko-Varga, a co-chair of the C-HPP project, told ProteoMonitor that while the chromosome-centric approach had been key to moving the HPP effort forward, a disease biology component would probably be necessary to obtain funding for the work.
"It's like the chicken and the egg," he said. "In order to get funding, you need to get going with your plans, but you also need the disease biology link. So I think the way the [C-HPP] will develop is that the chromosome groups will need that [disease-based] component as well."
While this might suggest that the two initiatives would perhaps eventually converge, Marko-Varga said the B/D-HPP was intended "more to focus on a few diseases and try to find out the details of mechanistic changes" as opposed to the broader aims of the C-HPP to characterize the entire human proteome on a chromosome-by-chromosome basis.
The two efforts will be collaborative, though, he said, especially in areas like obtaining access to "good, high-quality patient material," which he called a key "bottleneck" facing both projects.
[ pagebreak ]
As Aebersold noted, the HPP projects are supported by "three pillars" consisting of informatics resources like the UniProt and NextProt databases, the antibody-based resources of the Human Antibody Initiative, and the SRMAtlas compiled by the Institute for Systems Biology and ETH Zurich.
Discussion of these resources, too, demonstrated HUPO's move towards an emphasis on specific questions of disease and biology.
For instance, one manifestation of this emphasis was the launch several weeks ago of NextProt by the Swiss Institute of Bioinformatics and bioinformatics firm GeneBio (PM 8/5/2011). That resource, which builds upon the contents of the SIB-developed UniProtKB/Swiss-Prot database, aims to be a comprehensive source for human protein data, integrating a number of proteomics and protein research repositories, Amos Bairoch, professor of bioinformatics at the University of Geneva and leader of the project, told ProteoMonitor upon its release.
While UniProt aims for breadth – offering sequence and functional information on proteins from a number of different species – NextProt aims for depth: It contains information on human proteins only, but a considerably larger amount of such information, including data on identified peptides, antibodies, expression, and protein-protein interaction. It will also include tools and information aimed at the pharmaceutical industry.
Likewise, the SRMAtlas is undergoing an expansion to increase its relevance to specific research questions. At this week's meeting, ISB researcher Rob Moritz, one of the leaders of the project, presented an addition to the database called Passel that is meant to allow researchers to upload data on SRM assays performed in particular biological systems, he said.
At last year's HUPO conference, the ISB and ETH announced that they had completed an initial draft of the atlas, detailing more than 170,000 single reaction-monitoring assays – one each for at least five proteotypic peptides for each of the 20,300 human genes currently annotated as protein-encoding. This offers a set of pre-built, standardized SRM assays that scientists could use to more easily investigate proteins of interest (PM 9/24/2010).
However, Aebersold, a co-leader on the SRMAtlas project, noted this week that the assays in the atlas won't necessarily be best-suited to all biological sources and situations. "Even though [these assays] generate very good fragment ion spectra, they may not be detectable in a biology source for many reasons," he said.
Indeed, several researchers doing independent SRM-MS work have told ProteoMonitor that the assays contained in the SRMAtlas have not turned out to be the best ones for their projects. For example, in a discussion of his research using SRM-MS for drug-metabolism studies, Tohoku University scientist Tetsuya Terasaki said, "frankly, we are not too satisfied with those databases. We've compared results [using peptides] selected by our algorithm and [peptides] from the [SRMAtlas] and our experience is that we prefer to use our own selection criteria."
Passel will enable researchers to compile "multiple views of the same peptide by different groups," Moritz said, "and we can start to analyze how well does a particular peptide perform in the different biological experiments that people are doing."
According to ISB researcher Eric Deutsch, who is leading the Passel effort, the database "won't just be using published [SRM] transition lists from literature, but ideally we'll be using the chromatograms that have been submitted [by researchers]."
"So as you dial in a protein or proteins in SRMAtlas, you'll be able to see successful experiments that have been used [for that protein] before," he said.
Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.