Potential Oral Microbial Markers for Differential Diagnosis of Crohn’s Disease and Ulcerative Colitis Using Machine Learning Models

Although gut microbiome dysbiosis has been associated with inflammatory bowel disease (IBD), the relationship between oral microbiota and IBD remains poorly understood. This study aimed to identify unique patterns in saliva from patients explore potential microbial markers for differentiating Crohn’s (CD) ulcerative colitis (UC). A prospective cohort recruited (UC: n = 175, CD: 127) healthy controls (HC: 100) analyze their using 16S rRNA gene sequencing. Machine learning models (sparse partial least squares discriminant analysis (sPLS-DA)) were trained sequencing data classify CD UC. Taxonomic classification resulted 4041 phylotypes Kraken2 SILVA reference database. After quality filtering, 398 samples 124, HC: 99) 2711 included. Alpha diversity revealed significantly reduced richness of compared controls. The sPLS-DA model achieved high accuracy (mean accuracy: 0.908, AUC: 0.966) distinguishing vs. HC, as well good (0.846) AUC (0.923) These findings highlight distinct provide insights into diagnostic markers.