Automated Interpretation of Peptide Tandem Mass Spectra and Homology Searching

Abstract

Computer programs that match tandem mass spectrometry (MS/MS) data to database sequences typically search first for exact matches between the observed peptide mass and the mass calculated from the database (Eng et al. 1994). If exact matches cannot be found, some of these programs have the ability to find database sequences that do not match the observed peptide mass. The program PeptideSearch (Mann and Wilm 1994) uses the concept of a “peptide-sequence tag”, which is a partial sequence containing an unse-quenced mass on the C- and N-terminal ends of a short stretch of sequence. For triple-quadrupole and quadrupole/time-of-flight (QTOF) data of tryptic peptides, the short stretch of sequence is often easily determined from the prominent series of y-type ions that usually is present at masses greater than those of multiply charged precursor ions. Sequence tags are more difficult to obtain from ion-trap data, where high-mass b- and y-type ions are both typically present. Since peptide-sequence tags contain three parts — the N-terminal unsequenced mass, the C-terminal unsequenced mass and the intervening short sequence — it is possible to search a sequence database, requiring an exact match for only two of the three regions. Of course, by reducing the search constraints in this manner, the program is much more likely to find false positives, and it has been the author’s experience that several tens of possibilities are obtained.KeywordsTryptic PeptideTandem Mass SpectrumSequence CandidateDatabase MatchBiomed EnvironThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.