The problem with peptide presumption and low Mascot scoring

Abstract

Mascot, a database-search algorithm, is used to deduce an amino acid sequence from a peptide tandem mass spectrum. The magnitude of the Ions score associated with each peptide mostly reflects the extent of b-y ion matching in a collision-induced dissociation spectrum. Recently, several studies have reported peptides identified with abnormally low Ions scores. While a majority of the spectra in these studies may be correctly assigned, low-scoring spectra could lack discernible b-y ion fragments needed to clearly delineate a peptide sequence. It appears that low-scoring identification may be predicated primarily on judgmental parent ion mass accuracy and that justification to include such low-scoring peptides may be based on inaccurate false discovery rate modeling. It is likely that additional scientific experimentation is needed or appropriate methodologies adopted before substandard fragment ion matching can be considered proof of peptide identification.