International collaboration reveals common problems in mass spectrometry-based proteomics

Montreal, QC – New findings from an international collaboration, involving McGill University, the Research Institute of the McGill University Health Centre (MUHC) and the Human Proteome Organisation (HUPO) have been published a study in Nature Methods that shows common problems in mass spectrometry-based proteomics.

“Proteomics is the field that singles out the few significant proteins from the hundreds that may be present in a diagnostic sample,” says Dr Tommy Nilsson, co-author and researcher at the Research Institute of the MUHC and of McGill University. “It is important to associate the correct proteins with the correct condition. This process is incredibly complex. The aim of our study was to benchmark current analysis techniques worldwide and to identify potential bottlenecks.”

Twenty-seven labs worldwide were sent a standard sample of proteins to analyze using their usual techniques. Only seven of the 27 participating labs were accurate in detecting all the proteins and in the more challenging part of the study, only one lab succeeded. However, further analysis of their raw data, showed that all the proteins had been initially detected by all the labs involved but they had been rejected in later analyses.

“Our centralized analysis showed us the problems encountered while conducting this type of testing,” says Dr John Bergeron, senior author from McGill University and HUPO. “We found that a major contributing factor to erroneous reporting is at the database level. We expect once databases and search engines improve, the accuracy of reporting will as well.”

The goal of proteomics is to characterize all the proteins that are encoded from human DNA, similar to how all genes were identified as a result of the Human Genome Project. It is expected that proteomics will accelerate the identification of cause of many human diseases and that improved diagnosis and therapy will emerge using proteomic techniques.

“The new technology described in our paper will potentially enable clinicians to determine the causes of disease,” adds Dr. Bergeron.

The study, entitled “HUPO test sample study reveals common problems in mass spectrometry-based proteomics”, was authored by Alexander Bell (McGill University), Eric Deutsch (Research Institute, MUHC), Catherine Au (McGill University), Robert Kearney (CODA Genomics), Ron Beavis (BioGrammatics), Salvatore Sechi (NIDDK (NIH)), Tommy Nilsson (Research Institute, MUHC0, John Bergeron (McGill University) and the HUPO Test Sample Working Group.