One of the primary difficulties with protein identification using database search algorithms is the determination of post-translational modifications. When the appropriate modification isn't chosen in the initial search, the modification may not only be missed but this spectrum can be assigned to the wrong peptide producing a false positive. Sample handling alone can produce several different peptide modifications, most notably deamidation, oxidation, carbamylation, and improper alkylation. This work examines a new search mode designed to identify peptides with unexpected modifications. A trypsin digest of a HeLa cell protein lysate was fractionated by pI then analyzed using microfluidic-based nano-LC coupled to a Q-TOF mass spectrometer. Database searches were performed against the human IPI database and the initial database search was performed looking for only unmodified matches. Unexpected modifications were subsequently identified using an unassigned single mass gap mode which identifies any potential peptide modification by the mass of the modification. To determine the impact of MS/MS mass accuracy, searches were performed using the unassigned single mass gap homology search mode holding the precursor mass accuracy at 10 ppm while varying the fragment mass tolerances from 40 ppm to 1000 ppm. Results demonstrate that the combination of high mass accuracy for both precursor and product ion spectra and a new database search mode where modifications are not known beforehand, confident identification can be made for peptides that would previously have been missed.
Read full abstract