Algorithm for accurate similarity measurements of peptide mass fingerprints and its application.

We present a simple algorithm which allows accurate estimates of the similarity between peptide fingerprint mass spectra from matrix assisted laser desorption/ionization (MALDI) spectrometers. The algorithm, which is a combination of mass correlation and intensity rank correlation, was used to cluster similar spectra and to generate consensus spectra from a data store of more than 100,000 spectra. The resulting first spectra library of 1248 unambiguously identified different protein digests was used to search for missed cleavage patterns that have not been reported so far and to shed light on some peptide ionization characteristics. The findings of this study could be directly implemented in peptide mass fingerprint search algorithms to decrease the false positive error rate to <0.25%. Furthermore, the results contribute to the understanding of the peptide ionization process in MALDI experiments.