Evaluating DIA LiP-MS Analysis Workflows with a Hybrid LiP Proteome Benchmark

Emerging as a powerful structural proteomics approach, limited proteolysis mass spectrometry (LiP-MS) has been widely employed to interrogate proteome-wide protein structural alterations, identify drug targets and drug-binding pockets, and probe protein–protein interactions. However, LiP-MS-based proteomics data analysis is fundamentally different from that of conventional proteomics informatics. LiP-MS relies on peptide-centric analysis in order to pinpoint structural regions or residues within a protein that exhibit conformational changes. The presence of a large number of semitryptic peptides substantially increases LiP-MS data complexity. Moreover, there is a lack of consensus on the statistical criteria for defining structural changes. To evaluate informatics workflows for DIA-based LiP-MS, we generated a high-quality benchmark data set comprising more than 170,000 LiP peptides with defined composition. We then performed a comprehensive assessment of major DIA analysis platforms incorporating different spectral libraries, and introduced DIA-LiPQuan, an informatics pipeline tailored to DIA LiP-MS quantification and downstream analysis. Data reanalysis by DIA-LiPQuan with in silico libraries allows sensitive and robust detection of both site-specific structural remodeling of proteins and drug-bound protein targets from the cellular proteome. Collectively, our study provides a valuable benchmark resource and informatics package for LiP-MS data mining, which would facilitate its broader applications in structural proteomics and drug discovery.