A Unified Framework for Batch Correction and Imputation in Modern Mass Spectrometry Proteomics

NMFBatch reduces technical variation by effectively correcting batch effects in proteomics data, demonstrated across diverse datasets including plasma and single-cell samples.

• Why it matters: Batch effects, inter-laboratory variation, and signal drift hinder accurate biological interpretation in mass spectrometry proteomics, especially in single-cell applications.
• What they did: The method combines non-negative matrix factorization with generalized additive modeling to jointly address discrete and continuous technical effects, handling missing data through masking and imputation, validated on multiple datasets.
• The result: NMFBatch consistently decreases technical variation while maintaining biological signals, establishing a robust framework for batch correction, harmonization, and integrative analysis in proteomics research.