MetaDIA: A Novel Database Reduction Strategy for DIA Human Gut Metaproteomics

Abstract Background Microbiomes, especially within the gut, are complex and may comprise hundreds of species. The identification of peptides in metaproteomics presents a significant challenge, as it involves matching peptides to mass spectra within an enormous search space for complex and unknown samples. This poses difficulties for both the accuracy and the speed of identification. Specifically, analysis of data-independent acquisition (DIA) datasets has relied on libraries constructed from prior data-dependent acquisition (DDA) results. This approach requires running the samples in DDA mode to construct a library from the identified results, which can then be used for the DIA data. However, this method is resource-intensive, consumes samples, and limits identification to peptides previously identified by DDA. These limitations restrict the application of DIA in metaproteomics research. Results We introduced a novel strategy to reduce the search space by utilizing species abundance and functional abundance information from the microbiome to score each peptide and prioritize those most likely to be detected. Employing this strategy, we have developed and optimized a workflow called MetaDIA for analysis of microbiome DIA data, which operates independently of DDA assistance. Our method demonstrated strong consistency with the traditional DDA-based library approach at both protein and functional levels. Conclusion Our approach successfully created a smaller, yet sufficient database for DIA data search requirements in metaproteomics, showing high consistency with results from the conventional DDA-based library. We believe this method can facilitate the application of DIA in metaproteomics.