|
|
Institute
of Computing Technology,
Chinese
Academy of Sciences,
Beijing, China
|
Overview
Accurate proteome-wide quantitation is a great challenge for mass spectrometry based proteomics and its central component is informatics. We deeply investigated current quantitation software programs involving full MS scan based isotope labeling quantification strategies, and found that accuracy can be further improved and more consideration should be given to accuracy evaluation. Aiming to solving these crucial problems, we developed a software program named pQuant, integrating combinational, regression, clustering and hypothesis testing algorithms, and made detailed comparison with two of the most popular quantitation programs, i.e. Census and MaxQuant, using large-scale datasets. The results showed that over 38.2% of the quantified peptides had interfered signals, which accounted for more than half of all the signals of each peptide. In detecting these interfered signals and obtaining accurate quantitation results, pQuant was observably superior to Census and, if not better than, at least not inferior to MaxQuant. As for the evaluation of peptide ratio accuracy, the ERV (Estimated Ratio Variation) of pQuant was more trustworthy than the determinant score R2 of Census, and MaxQuant totally lacks this part. More importantly, evaluation of the accuracy of protein ratios was provided in pQuant but sorely lacking in the other two programs.
Download
Download pQuant(Sorry, she is not ready yet)
Download Test Data:
0425-HUMAN-SILAC-1to1-08.RAW
|
|