Retention time (RT) alignment is crucial for LC-MS-based proteomic and metabolomic analyses, especially in large cohorts. Existing tools struggle with simultaneous monotonic and non-monotonic RT shifts. This paper introduces DeepRTAlign, a deep learning-based RT alignment tool. DeepRTAlign shows improved performance against current state-of-the-art approaches on various datasets, enhancing identification sensitivity without compromising quantitative accuracy. Its application in predicting hepatocellular carcinoma recurrence demonstrates its utility in downstream biological analyses.
