MassMatrix

MassMatrix matches tandem mass spectra to theoretical peptide sequences from protein databases to identify and quantify peptides for proteomics analyses.

Key Features:

• Probabilistic Scoring Model: Uses a mass accuracy-sensitive probabilistic approach to evaluate and rank peptide-spectrum matches.
• High Sensitivity and Specificity: Demonstrates superior sensitivity relative to MASCOT, SEQUEST, X!Tandem, and OMSSA at a given specificity while maintaining a low false positive rate (~2%).
• Unique Peptide/Spectrum Matches: Manually validated true positives correspond to unique peptide/spectrum matches.
• Robustness to Decoys and PTMs: Performance in high mass accuracy searches is not significantly affected by the presence of decoy sequences or additional variable post-translational modifications.
• Support for Multiple Variable Modifications: Handles complex datasets that include multiple variable modifications.
• Efficient Processing: Achieves improved search time efficiency compared with counterpart MS/MS search software.
• High Throughput on Distributed Systems: When deployed on distributed memory clusters, can process approximately 100,000 spectra per hour against a comprehensive human database with eight variable modifications.

Scientific Applications:

• Peptide Identification and Quantitation: Identification and quantitation of peptides in proteomics experiments using tandem mass spectrometry data.
• High Mass-Accuracy MS/MS Analyses: Analysis of high mass accuracy MS/MS datasets, including searches for post-translational modifications.
• Benchmarking of MS/MS Search Algorithms: Comparative benchmarking against other MS/MS search tools such as MASCOT, SEQUEST, X!Tandem, and OMSSA.

Methodology:

Matches tandem mass spectra to theoretical peptide sequences derived from protein databases using a mass accuracy-sensitive probabilistic scoring model; supports searches with multiple variable post-translational modifications; evaluated on high mass accuracy datasets and benchmarked against MASCOT, SEQUEST, X!Tandem, and OMSSA, and executed on distributed memory clusters for high-throughput processing (~100,000 spectra/hour against a human database with eight variable modifications).