Fig. 3From: Development of a 32-gene signature using machine learning for accurate prediction of inflammatory bowel diseaseThe 32-gene signature shows similar expression patterns in all cohorts and are mainly related to immune response. A iHMP, (B) GSE75214, (C) GSE3365, (D) GSE10616, (E) GSE6731 (F) GSE1152, (G) GSE112366, (H) GSE53306, (I) GSE38713, (J) GSE9452, and (K) GSE22619. Row Z-score gene expression heatmaps were generated using Log2(TPM + 1) values of the iHMP cohort and other cohorts' microarray expression profiles. L Gene-ontology analysis using metascape (http://metascape.org) of the 32-gene signature. The red dashed lines indicate upregulated genesBack to article page