ABSTRACT
Peptide sequence identification using tandem mass spectroscopy remains a major challenge for complex proteomic studies. Peptide matching algorithms require the accurate determination of both the mass and charge of the precursor ion and accommodate uncertainties in these properties by using a wide precursor mass tolerance and by testing, for each spectrum, several possible candidate charges. Using a data acquisition strategy that includes obtaining narrow mass-range MS(1) "zoom" scans, we describe here a post-acquisition algorithm dubbed mass and charge (Z) inference engine (MAZIE), which accurately determines the charge and monoisotopic mass of precursor ions on a low-resolution Thermo LTQ-XL mass spectrometer. This is achieved by examining the isotopic distribution obtained in the preceding MS(1) zoom spectrum and comparing to theoretical distributions for candidate charge states from +1 to +4. MAZIE then writes modified data files with the corrected monoisotopic mass and charge. We have validated MAZIE results by comparing the sequence search results obtained with the MAZIE-generated data files to results using the unmodified data files. Using two different search algorithms and a false discovery rate filter, we found that MAZIE-interpreted data resulted in 80% (using SEQUEST) and 30% (using OMSSA) more high-confidence sequence identifications. Analyses of these results indicate that the accurate determination of the precursor ion mass greatly facilitates the ability to differentiate between true and false positive matches, while the determination of the precursor ion charge reduces the overall search time but does not significantly reduce the ambiguity of interpreting the search results. MAZIE is distributed as an open-source PERL script.
