Application of SWATH Mass Spectrometry and Machine Learning in Diagnosis of Inflammatory Bowel Disease Based on Stool Proteome

Our hypothesis was that conducting proteomic analysis on clinical laboratory samples which are intended for the fecal calprotectin test would enable the develop-ment of a highly sensitive and specific non-invasive stool test based on mass spec-trometry. To investigate this hypothesis, we combined and applied our expertise in basic research, clinical practice, and bioinformatics to develop a precise ma-chine-learning model for the accurate diagnosis of active IBD from symptomatic non-IBDpatients.