A Perl procedure for protein identification by Peptide Mass Fingerprinting

Area of Science:

• Proteomics
• Bioinformatics
• Mass Spectrometry

Background:

• Protein identification is crucial in proteomics.
• Peptide Mass Fingerprinting (PMF) is a common approach using mass spectrometry (MS) data and database searching.
• Existing scoring and statistical evaluation methods require improvement.

Purpose of the Study:

• To present MsPI (Mass spectrometry Protein Identification), a Perl-based tool for protein identification via PMF.
• To enhance scoring methods and statistical evaluation of results.
• To provide a tool that removes contaminant masses and assigns p-values to candidate proteins.

Main Methods:

• Development of a Perl procedure (MsPI) for PMF.
• Implementation of scoring methods from existing literature.
• Inclusion of a strategy for contaminant mass removal.
• Integration of a statistical method for p-value assignment.
• Comparison of MsPI with Piums and Mascot on a dataset of 10 protein samples.

Main Results:

• MsPI demonstrated superior performance compared to Piums and Mascot.
• MsPI correctly identified "true" proteins in 9 out of 10 samples.
• MsPI generated shorter candidate lists, reducing false positives compared to Mascot.
• Piums identified "true" proteins in only 4 out of 10 samples.

Conclusions:

• MsPI offers improved accuracy and reduced false positives in protein identification.
• The Perl-based nature of MsPI allows for user extensibility and customization.
• MsPI represents a valuable advancement in PMF-based protein identification tools.