Achieving quantitative reproducibility in label-free multisite DIA experiments through multirun alignment.

Area of Science:

• Proteomics
• Analytical Chemistry
• Bioinformatics

Background:

• Quantitative proteomics using Data-Independent Acquisition (DIA) is widely used but hampered by inconsistent results across multiple LC-MS/MS instruments.
• Retention time shifts between runs from different sites cause quantification errors and data loss, limiting parallel data acquisition.

Purpose of the Study:

• To develop and evaluate novel multirun chromatogram alignment strategies for improved quantitative proteomics.
• To address the bottleneck of inconsistent quantification in multi-instrument, multi-site DIA studies.

Main Methods:

• Implementation of three multirun chromatogram alignment strategies: reference-based Star method, and novel reference-free MST and Progressive alignment.
• Application of these strategies within the DIAlignR workflow to diverse datasets, including gold-standard, multi-species, and large-scale plasma samples.

Main Results:

• Multirun alignment reduced quantitation error-rate by threefold compared to non-aligned data.
• MST alignment decreased cross-site coefficient of variation (CV) by 50% for abundant proteins.
• Reanalysis of 949 plasma runs identified over 50% more proteins related to insulin resistance and viral infections.

Conclusions:

• Reference-free alignment strategies (MST, Progressive) provide quantitatively accurate data matrices for heterogeneous LC-MS/MS studies.
• DIAlignR's multirun alignment enhances precision, controls quantitative error, and increases proteome coverage.
• These methods are crucial for robust and scalable quantitative proteomics in multi-center collaborations.