Probability-Based Protein Identification for Post-Translation Modifications and Amino Acid Variants Using Peptide Mass Fingerprint Data

Methods: We apply a log-likelihood ratio calculation to determine the probability that a protein is present within the sample. The model distinguishes the null hypothesis: a set of peaks in the spectrum is generated by the random background, from the alternative hypothesis: the same set of peaks is generated by peptides corresponding to a specific protein. Peak assignment is based on the log-likelihood ratio as oppose to matching peaks with peptides within the mass tolerance. The mass spectrum is first matched with unmodified protein sequences. The top scored proteins in the search result are then used to construct a database that includes PTMs and amino acid variants, against which the spectrum is searched again. Results are ranked by their probability scores.