FPTMS: Frequency-based approach to identify the peptide from the low-energy collision-induced dissociation tandem mass spectra.

The database search method is a widely accepted method to assign a peptide to the tandem mass spectra. In this study, a new flexible method- FPTMS is introduced to interpret the tandem mass spectra with the known peptide sequences in a protein database. Here the frequency of occurrence of fragment ion peaks extracted from the extensive spectral library is used to predict the theoretical tandem mass spectra of the peptides. The dot product scoring and windowed-xcorr scoring methods were implemented to score the experimental spectrum against the theoretical peptide spectra. Windowed-xcorr is introduced to tackle the mass errors and the cleavage position of the fragmentation process. The new method with windowed-xcorr shows an improved identification rate compared to the existing search engines Crux-Tide and X!Tandem at 1% False Discovery Rate (FDR) for the dataset considered in this study. Identifying or sequencing of the peptide from tandem mass spectra is an important application in mass spectrometry-based proteomics. Collision-induced dissociation (CID) fragmentation spectra have been widely used to develop a peptide identification algorithm using database search strategy. CID fragmentation behavior is a complex process and found to have dependency on the sequences of peptide, charge state, and residue content. The inclusion of more features of peptide fragmentation behavior and adaptable scoring algorithm improves the efficiency of the peptide identification algorithm. Unlabelled Image • FPTMS identifies the peptide from the low energy CID spectrum of tryptic peptides. • The relative frequency of occurrence of fragment ion feature is used to predict the theoretical spectrum of the peptide reliably. • Sliding window cross-correlation is used for scoring the experimental spectrum against the candidate peptide spectrum. • FPTMS(windowed-xcorr) shows an improved identification rate compared to existing algorithms: Crux-Tide & X!Tandem at 1% FDR. [ABSTRACT FROM AUTHOR]