ChimST: An Efficient Spectral Library Search Tool for Peptide Identification from Chimeric Spectra in Data-Dependent...

Area of Science:

• Computational proteomics
• Bioinformatics
• Mass spectrometry data analysis

Background:

• Accurate peptide identification from MS/MS spectra is crucial but challenging in computational shotgun proteomics.
• Existing spectral library search tools struggle with chimeric spectra, where multiple peptides co-fragment.

Purpose of the Study:

• To develop a novel computational tool, ChimST, for reliable identification of multiple peptides from chimeric spectra.
• To enhance the performance of spectral library searching in proteomics.

Main Methods:

• ChimST associates MS/MS spectra with precursor features.
• It considers peptide candidates from a spectral library for each feature.
• A scoring system evaluates collective peptide interpretations of the query spectrum.

Main Results:

• ChimST significantly outperforms state-of-the-art tools (SpectraST, reSpect, MSPLIT).
• It achieves higher numbers of peptide-spectrum matches and unique peptides identified.
• Performance gains are particularly notable with broad acquisition isolation windows.

Conclusions:

• ChimST offers a robust solution for identifying multiple peptides in complex chimeric spectra.
• This tool advances computational proteomics by improving the analysis of challenging MS/MS data.
• ChimST enhances the sensitivity and accuracy of peptide identification in shotgun proteomics.