Deep plasma proteomics with data-independent acquisition: A fastlane towards biomarkers identification

Plasma proteomic is a precious tool in human disease research, but requires extensive sample preparation in order to perform in-depth analysis and biomarker discovery using traditional Data-Dependent Acquisition (DDA). Here, we highlight the efficacy of combining moderate plasma prefractionation and Data-Independent Acquisition (DIA) to significantly improve proteome coverage and depth, while remaining cost- and time-efficient. Using human plasma collected from a 20-patient COVID-19 cohort, our method utilises commonly available solutions for depletion, sample preparation, and fractionation, followed by 3 LC-MS/MS injections for a 360-minutes DIA run time. DIA-NN software was then used for precursor identification, and the QFeatures R package was used for protein aggregation. We detect 1,321 proteins on average per patient, and 2,031 unique proteins across the cohort. Filtering precursors present in under 25% of patients, we still detect 1,230 average proteins and 1,590 unique proteins, indicating robust protein identification. Differential analysis further demonstrates the applicability of this method for plasma proteomic research and clinical biomarker identification. In summary, this study introduces a streamlined, cost- and time-effective approach to deep plasma proteome analysis, expanding its utility beyond classical research environments and enabling larger-scale multi-omics investigations in clinical settings.