Abstract
In proteomics, identifying proteins expressed in a biological sample requires interpreting fragmentation spectra obtained from their peptides. This can be achieved by matching experimental fragmentation spectra with theoretical spectra generated in silico from a protein sequence database. Often, the reference database comes from a sequenced genotype different from the one under investigation. In such cases, non-synonymous genetic polymorphisms can introduce amino acid variations, leading to discrepancies between the theoretical and experimental peptide masses, which may hinder spectrum interpretation.Open Modification Search (OMS) approaches help to overcome this constraint.
In this work we used SpecOMS (David et al. 2017) on a gold standard dataset produced from four maize lines. Each line has its own reference database, containing intra-specific genetic variations. SpecOMS does not impose a mass filter on theoretical spectra and we demonstrate here that it allows for the correct interpretation of a PSM. By comparing the results of the classical mass matching engines for a selected maize line on the correct protein database (called self-interrogation) and the SpecOMS results using a different protein database (called cross-interrogation), we show that SpecOMS assigns correctly the PSM to the peptide sequence with the amino acid variation and is also able to spot correctly the position of the mass delta.