Protein identification algorithms

J. Eriksson and D. Fenyo. Probity A Protein Identification Algorithm with Accurate Assignment of the Statistical Significance of the Results./. Proteome Res., 3, no. 1 (2004) 32-36. 

CRITICAL ASSESSMENT OF THE METHOD Since MS instrumentations today are robust and sensitive, protein identification by PMF is quite good. The rapid development of protein and gene sequence databases provides an ever more consistent platform for searches. Furthermore, improvements in search algorithms and the application of sophisticated statistical methods has improved the reliability... 

Routinely use peptide mass database algorithms for protein identification 39 26%... 

A complementary approach to protein identification is based on peptide seqnencing analysis (PSA) by collision-induced dissociation (CID) in MS-MS. Por proteomics studies, the de novo sequencing of MS-MS spectra is too time-consuming. Therefore, algorithms have been developed to either provide antomated MS-MS spectmm interpretation, or perform protein identification by means of a SEQUEST database search [6-7]. In the latter case, the precursor m/z values are matched against a virtual digestion of all the proteins in the database. Sequence ions are predicted for the peptides that match the precnrsor m/z values, and compared with the measured MS-MS data. A correlation score is calculated for each match. 

The SEQUEST database search algorithm (Ch. 17.6.2) is a powerful tool in protein identification, although like any statistical tool it relies on the jndgments of the user to critically evaluate its results. Problems may especially arise, when unanticipated modifications are present in the protein investigated. 

---