Optimizing Spectral Reference Library Generation Strategies for Data-Independent Acquisition Proteomics in Clinical Cancer Cohorts
Data-independent acquisition (DIA)-mass spectrometry (MS) data is typically searched using spectral reference libraries (SRLs) created from data-dependent acquisition (DDA). Shortcomings with this approach include inherent missingness, low reproducibility, and bias to high abundant peptides. We systematically compared the performance of four SRL generation strategies: (1) fractionated DDA, (2) fractionated DIA, (3) in silico SRL derived library refined using unfractionated sample DIA files (SRL-free DIA), and (4) gas phase fractionation (GPF) DIA on 9 cancer cohorts accounting for 246 human samples. Each was evaluated in terms of library size, quantitated peptides, q-value scoring, missingness, peptide intensity, and differentially abundant (DA) proteins. In a prostate adenocarcinoma set, fractionated DDA displayed the highest number of identifications, while SRL-free DIA had the lowest number of missing values (∼40%). Peptides uniquely quantified by SRL-free DIA showed favorable within-search scoring characteristics, higher intensity values (>50% improvement), and the lowest missingness (14% improvement) leading to more significantly reported DA proteins. Fractionated DDA SRLs performed less favorably in most of the metrics tested, challenging the assumption that larger SRLs improve high-quality quantitation. SRL-free DIA closely represents the samples under study and offers a higher quality alternative for large-scale clinical proteomic analysis where sample availability is limited.
Authors
- Peter G. Hains (ORCID: https://orcid.org/0000-0002-7276-1760)
- Clare Loudon (ORCID: https://orcid.org/0000-0002-5378-4297)
- Phillip James Robinson (ORCID: https://orcid.org/0000-0002-7878-0313)
- Natasha Lucas (ORCID: https://orcid.org/0000-0002-7656-6866)
- Steven G. Williams (ORCID: https://orcid.org/0000-0003-4647-436X)
- Daniel Bucio-Noble (ORCID: https://orcid.org/0000-0001-7603-8732)
- Jennifer M.S. Koh (ORCID: https://orcid.org/0000-0003-2140-3151)
- Daniela-Lee Smith (ORCID: https://orcid.org/0000-0002-1410-5079)
- Dylan Xavier (ORCID: https://orcid.org/0000-0003-2601-9343)
- Erin K. Sykes (ORCID: https://orcid.org/0000-0002-1582-5238)
- Erin M. Humphries (ORCID: https://orcid.org/0000-0001-7186-0620)
Institutions
- The University of Sydney (AU)
Publication Details
- Journal
- Journal of Proteome Research
- Published
- 2026-11-18
- DOI
- https://doi.org/10.1021/acs.jproteome.5c01025
- Primary Topic
- Advanced Proteomics Techniques and Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Australian Cancer Research Foundation
- National Breast Cancer Foundation
- Department of Health and Ageing, Australian Government
- Ian Potter Foundation
- Cancer Institute NSW
- University of Sydney
- NSW Health
- National Health and Medical Research Council
- Cancer Council NSW