HUPO Plasma Proteome Project 2007 Workshop Report

Abstract

The HUPO Plasma Proteome Project (PPP) held its annual workshop at the 6th HUPO World Congress in Seoul, Korea, October 7, 2007. The agenda was organized around the aims of the next phase of the PPP. The pilot phase accomplished its goals of providing reference specimens of serum and plasma to a large international network of collaborating laboratories, comparing protein identifications with various fractionation and analytical technology platforms, introducing new technologies from vendor partners, generating extensive submissions to open source databases (www.peptideatlas. org/hupo/hppp (ISB); www.ebi.ac.uk/pride (EBI)) and bioinformatics insights, and laying a foundation for integration of plasma studies with organ proteome and disease biomarker projects. Results were published in a special issue of Proteomics (1), a Wiley book (2), and a special analysis in Nature Biotechnology (3). The specimens and databases have been utilized for many other studies. Participants at the PPP workshop in Long Beach, California, in November 2006 agreed to develop the current plans. The 2007 workshop was organized with panels of leading scientists in proteomics describing their commitments. The overriding principle is the sharing of data to advance the field and facilitate much needed comparative analyses. The first panel comprised laboratory leaders with advanced technology platforms. Ruedi Aebersold described the N-glycosite Nx(S/T) proteotypic peptide library approach for targeted analyses. Christie Hunter presented applications of multiple reaction monitoring with multiple transitions for each labeled peptide. Denis Hochstrasser emphasized the preanalytical variables and the advantages of starting from tissue, followed by the multiplex analyses of plasma (or serum, if immunoassays). Bill Hancock introduced multilectin affinity chromatography for glycoproteins, including glycan characterization. Ole Jensen introduced microparticles and phosphoprotein analyses. Richard Smith explained powerful ion mobility separation mass spectrometry and electrospray emitter design. And Rong Zeng presented label-free, integrated analyses after ethanol precipitation and pH elution of plasma proteins from diabetics. The second panel presented highly complementary bioinformatics and database resources already impressively connected with the PPP. Henry Lam announced that PeptideAtlas at the Institute for Systems Biology now has a human PPPspecific subatlas (see above); as the total resource has grown, more and more of the newly reported multiobserved peptides are confirmatory. PeptideAtlas applies to submitted datasets its uniform data analysis pipeline, including MS/MS spectral library searches with SpectraST (joint with the National Institute of Standards and Technology), which are much faster than Sequest. Lennart Martens described typical pitfalls in the analysis of mass spectrometry data obtained from serum or plasma samples along with tools that help resolve these problems. He reported that the PPP datasets were the first external data submitted to PRIDE at the European Bioinformatics Institute (see above); PRIDE assists data dissemination and links to EBI resources. Ron Beavis noted that the PPP was also one of the first datasets in the Global Proteome Machine, GPMdb (with the distinctive prefix 101). Martin McIntosh emphasized statistical variables, starting with variability among individuals and populations to be studied and discussed that follow-up analysis of the same individual minimizes genetic variation. Akhilesh Pandey described the Proteinpedia portal for sharing protein data and annotations. Both PRIDE and Proteinpedia use Tranche at ProteomeCommons.org, created by Philip Andrews of the University of Michigan, for storage of large volume raw datasets. Tranche provides encrypted file sharing among collaborators as well as open access datasets when made available. Contributors to the PPP should submit complex datasets with full experimental annotation specified by the HUPO Protein Standards. Users can find unaltered submissions at PRIDE, X!Tandem analyses at Global Proteome Machine, and TransProteomic Pipeline-processed datasets at PeptideAtlas. The final panel, moderated by John Bergeron as chair of HUPO Initiatives, demonstrated that the PPP and all other HUPO initiatives are complementary. EDTA-plasma specimens should be collected as part of organ-based studies for analysis in the same laboratories, followed by sharing of data on plasma and tissue findings with the PPP through the informatics resources just described. Helmut E. Meyer (Brain Proteome Project) reinforced the priority for starting from From ‡Human Genetics and Public Health, Center for Computational Medicine and Biology, University of Michigan, Ann Arbor, Michigan 48109, ¶Institute for Molecular Systems Biology, HPT E 76 ETH Honggerberg, CH-8093 ZAurich, Switzerland, and Department of Biochemistry, Yonsei Proteome Research Center & Biomedical Proteome Research Center, Room 423 Industry-University Building, Yonsei University, 134 Shinchon-dong, Sudaemoon-ku, Seoul, 120-749 Korea 1 The abbreviations used are: PPP, HUPO Plasma Proteome Project; PRIDE, Proteomics Identification Database. 2 This citation refers to just one of the articles used in this publication. HUPO Views