Potential Oral Microbial Markers for Differential Diagnosis of Crohn's Disease and Ulcerative Colitis Using Machine...

Area of Science:

• Microbiology
• Gastroenterology
• Genomics

Background:

• Gut microbiome dysbiosis is linked to inflammatory bowel disease (IBD).
• The role of oral microbiota in IBD pathogenesis and diagnosis is not well understood.
• Distinguishing between Crohn's disease (CD) and ulcerative colitis (UC) is clinically important.

Purpose of the Study:

• To identify unique oral microbiome patterns in IBD patients.
• To explore potential salivary microbial biomarkers for differentiating CD and UC.
• To assess the accuracy of machine learning models in classifying IBD subtypes.

Main Methods:

• Prospective cohort study including IBD patients (UC, CD) and healthy controls (HC).
• Oral microbiota analysis using 16S rRNA gene sequencing.
• Machine learning models (sPLS-DA) applied for classification and differentiation.

Main Results:

• Reduced microbial richness observed in IBD patients compared to HC.
• High accuracy achieved in distinguishing IBD from HC (accuracy: 0.908, AUC: 0.966).
• Good accuracy obtained in differentiating CD from UC (accuracy: 0.846, AUC: 0.923).

Conclusions:

• Distinct oral microbiome profiles are associated with IBD.
• Salivary microbial patterns show potential as diagnostic markers for IBD subtypes.
• Machine learning effectively utilizes oral microbiota data for IBD classification.