In Communications biology
Previous studies have conducted time course characterization of murine colitis models through transcriptional profiling of differential expression. We characterize the transcriptional landscape of acute and chronic models of dextran sodium sulfate (DSS) and adoptive transfer (AT) colitis to derive temporal gene expression and splicing signatures in blood and colonic tissue in order to capture dynamics of colitis remission and relapse. We identify sub networks of patient-derived causal networks that are enriched in these temporal signatures to distinguish acute and chronic disease components within the broader molecular landscape of IBD. The interaction between the DSS phenotype and chronological time-point naturally defines parsimonious temporal gene expression and splicing signatures associated with acute and chronic phases disease (as opposed to ordinary time-specific differential expression/splicing). We show these expression and splicing signatures are largely orthogonal, i.e. affect different genetic bodies, and that using machine learning, signatures are predictive of histopathological measures from both blood and intestinal data in murine colitis models as well as an independent cohort of IBD patients. Through access to longitudinal multi-scale profiling from disease tissue in IBD patient cohorts, we can apply this machine learning pipeline to generation of direct patient temporal multimodal regulatory signatures for prediction of histopathological outcomes.
Peters Lauren A, Friedman Joshua R, Stojmirovic Aleksandar, Hagen Jacob, Houten Sander, Dodatko Tetyana, Amaro Mariana P, Restrepo Paula, Chai Zhi, Rodrigo Mora J, Raymond Holly A, Curran Mark, Dobrin Radu, Das Anuk, Xiong Huabao, Schadt Eric E, Argmann Carmen, Losic Bojan
2023-Jan-24
