Study aims to improve common protein discovery method
(18 July 2009: VIDYYA MEDICAL NEWS SERVICE) -- A new study has identified several common problems that hinder the accuracy and reproducibility of using mass spectrometry to discover potential protein biomarkers of cancer. With feedback and training for laboratories conducting these analyses, the study authors believe, these issues can be overcome. The study was published in the June 2009 Nature Methods.
Conducted by the international Human Proteome Organization (HUPO), the study involved 27 laboratories that were asked to independently identify test samples spiked with 20 human proteins, each one with at least one specific peptide (a sequence of amino acids that join together to form proteins) of approximately the same mass. Each lab performed a mass spectrometry-based analysis of the samples, using their typical methodology, and reported the results and their raw data to HUPO for a centralized analysis.
Only 7 of the labs correctly identified all 20 proteins, and only 1 identified all of the peptides. However, when the raw data submitted by each lab was reanalyzed by the study leaders, they found that “members of each of the labs, with a few exceptions, had in fact generated mass spectrometry data of very high quality, more than sufficient to identify” all of the proteins and most of the peptides, the study authors wrote. Across the participating labs, several common problems led to the protein misidentifications, the researchers reported, such as discrepancies in searching the protein databases to match peptides to their respective proteins and test sample contamination.
Aside from the database issues, many of the problems encountered by individual labs “could be rapidly found and diagnosed,” explained the study's lead investigator, Dr. John Bergeron. In fact, after providing feedback to the participating labs on the problems discovered by the centralized analysis, all of the labs were then able to correctly identify the 20 proteins.
The study by HUPO “shows that with sufficient experimental care and technical training, currently available mass spectrometry-based methods” could produce accurate results, wrote Dr. Ruedi Aebersold, director of the Institute of Molecular Systems Biology at the University of Zurich in Switzerland, in an accompanying editorial. “It also makes a strong case for making data from proteomic studies accessible for comparison and meta-analyses and for the optimization and testing of software tools.”
Return to Vidyya Medical News Service for 18 July 2009
© Vidyya. All rights reserved.
Information appearing on the Vidyya Medical News Service is not intended as a substitute for professional medical care. Seek professional medical help and follow your health care provider's advice.
Interested in subscribing to our daily e-mail newsletter? Send an email to Vidyya@vidyya.com with the word subscribe in the subject field.