Error-tolerant identification of peptides in sequence databases by peptide sequence tags

We demonstrate a new approach to the identification of mass spectrometrically fragmented peptides. A fragmentation spectrum usually contains a short, easily identifiable series of sequence ions, which yields a partial sequence. This partial sequence divides the peptide into three parts-regions 1, 2, and 3-characterized by the added mass m1 of region 1, the partial sequence of region 2, and the added mass m3 of region 3. We call the construct, m1 partial sequence m3, a "peptide sequence tag" and show that it is a highly specific identifier of the peptide. An algorithm developed here that uses the sequence tag to find the peptide in a sequence database is up to 1 million-fold more discriminating than the partial sequence information alone. Peptides can be identified even in the presence of an unknown posttranslational modification or an amino acid substitution between an entry in the sequence database and the measured peptide. These concepts are demonstrated with model and practical examples of electrospray mass spectrometry/mass spectrometry of tryptic peptides. Just two to three amino acid residues derived by fragmentation are enough to identify these peptides. In peptide mapping applications, even less information is necessary.