Computational Optimization of Spectral Library Size Improves DIA-MS Proteome Coverage and Applications to 15 Tumors

Area of Science:

• Proteomics
• Mass Spectrometry
• Computational Biology

Background:

• Accurate peptide and protein identification from data-independent acquisition mass spectrometry (DIA-MS) relies on appropriately sized, project-specific spectral libraries.
• Existing methods often use large, general libraries that may not be optimal for specific datasets.

Purpose of the Study:

• To introduce subLib, a computational strategy for optimizing spectral library size for specific DIA-MS datasets.
• To improve the depth of proteome coverage in DIA-MS analyses.

Main Methods:

• subLib strategy involves preliminary analysis of DIA-MS data to inform spectral library optimization.
• A comprehensive spectral library is used as the basis for optimization.
• The strategy was tested on colorectal tumor samples and a larger cohort of carcinoma samples.

Main Results:

• subLib increased peptide precursor identifications by 41.2% and protein group identifications by 35.6% in a test dataset.
• Application to 389 carcinoma samples across 15 datasets yielded up to 39.2% increase in peptide precursors and 19.0% in protein groups.
• Demonstrated significant improvements in proteome coverage compared to a pan-human library approach.

Conclusions:

• The subLib strategy effectively optimizes spectral library size for DIA-MS data.
• This optimization significantly enhances proteome coverage and the number of identified peptides and proteins.
• subLib represents a valuable computational tool for improving DIA-MS data analysis and discovery.