Detection of Viral Proteins in Human Cells Lines by Xeno-Proteomics: Elimination of the Last Valid Excuse for Not Testing Every Cellular Proteome Dataset for Viral Proteins

Chernobrovkin, Alexey L.; Zubarev, Roman A.

Viral contamination differs from microbial contamination in terms of the proteome size: viral proteome can be represented by only a few proteins, while microbial proteome is much larger (. 500 proteins). Thus, detecting viral contamination is more challenging. Direct transfer of convenient data-analysis approaches widely used in single-organism proteomics may result in ''false alarm'' events [22, 23] . It should also be taken into account that some human proteins exhibit high degree of homology to retroviral proteins. In such case standard target-decoy strategy may fail [36] , so the special attention should be paid to the peptide-spectrum matches attributed to retroviral proteins. For instance, additional filtering based on delta score (score difference between the best and next best matches) could help reducing the amount of false positives. We also recommend manual rechecking of the MS/MS spectra and peptide assignments, as well as the uniqueness of the viral peptide sequences.