Improved FFPE Tissue Cancer-Associated Protein Detection using a Recombinant Protein Spectral Library (rPSL) DIA-MS Workflow

Formalin-fixed, paraffin-embedded (FFPE) tumour archives represent a vast, clinically annotated resource for biomarker discovery. However, reliable detection and quantification of lower-abundance proteins, particularly cancer biomarkers, remain challenging. Here, we applied an optimised recombinant protein spectral library data-independent acquisition MS (rPSL-DIA-MS) workflow to FFPE colorectal cancer (CRC) tissues to demonstrate efficient protein extraction, compare library-based and library-free DIA analysis using FragPipe/DIA-NN and Spectronaut, and assess sensitive detection and quantification of candidate biomarkers. Proteins were extracted from 20 CRC FFPE tumours and analysed by DIA-MS. An independent cohort (n = 20) processed separately was used to evaluate reproducibility. Optimised extraction and deep DDA library generation enabled identification of > 7,000 proteins (> 5,500 quantified) from FFPE CRC tissues. Across both platforms, library-based DIA achieved greater proteome coverage than library-free analysis. Recombinant proteins representing 34 cancer-associated proteins were used to generate rPSL-only and merged biological-rPSL libraries. rPSL-based DIA improved detection of most cancer-associated proteins, enabling quantification of all 34 proteins with multiple high-stringency peptides and increased coverage compared with standard biological-library and library-free DIA-MS workflows. Independent batch analysis showed minimal variation. Overall, rPSL-DIA-MS enables deep, sensitive and reproducible proteomic profiling of CRC FFPE tissues, supporting retrospective biomarker discovery in archival cohorts.