Integration of longitudinal quality metrics enhances differential analysis in noisy large-scale Mass Spectrometry(MS)-based proteomics experiments
Date:
This work presents a method for improving differential analysis in large-scale, noisy DIA-based proteomics experiments by integrating longitudinal quality metrics into an Isolation Forest anomaly detection framework. By identifying poor-quality peptide-spectrum matches using metrics like shape quality score and fragment noise, and incorporating these scores into weighted regression, the approach corrects low-quality and missing measurements, leading to more accurate and reliable detection of differential abundance compared to conventional methods.