Towards fast and reliable candidate peptide indexing for Tandem Mass Spectrometry data

Many algorithms for protein and peptide identification from Tandem Mass Spectrometry (MS/MS) data have been published. The majority of such methods are based on sequence or spectral database search: experimental spectra are matched to theoretical spectra characterizing known peptide sequences. Hits define candidate peptides. In this context, the development of reliable candidate peptide indexing and ranking techniques for accelerated pre-processing of large datasets still remains a challenge. The present work takes up this challenge and proposes a method based on a low-complexity hierarchical search approach suitable for differently charged experimental peptides. The technique efficiently utilizes the statistical information on intensity and mass-to-charge (m/z) values of the experimental peaks and rationally avoids a time-consuming, one-to-all-database comparison. A robust ranking methodology based on probabilistic likelihoods simplifies the interpretation of the obtained results. Outputting a short list of ranked candidate indices, the proposed indexing approach can be applied to pre-process experimental MS/MS spectra prior to submission to a search engine thus enhancing the quality of identification results.