Usefulness of Machine Learning-Based Gut Microbiome Analysis for Identifying Patients with Irritable Bowels Syndrome

Area of Science:

• Gastroenterology
• Microbiome Research
• Computational Biology

Background:

• Irritable Bowel Syndrome (IBS) diagnosis relies on subjective clinical symptoms.
• Objective diagnostic markers for IBS are needed.
• Gut microbiome alterations are implicated in IBS pathophysiology.

Purpose of the Study:

• To develop an objective prediction model for IBS using machine learning and gut microbiome analysis.
• To identify specific microbial and metabolic signatures associated with IBS.
• To validate the model's accuracy in distinguishing IBS patients from healthy controls.

Main Methods:

• Fecal samples and clinical data were collected from 85 IBS patients and 26 healthy controls.
• 16S ribosomal RNA sequencing analyzed gut microbiome profiles.
• Gas chromatography-mass spectrometry measured short-chain fatty acids.
• LASSO logistic regression and machine learning were employed for model development.

Main Results:

• IBS patients exhibited significantly lower fecal microbiome alpha-diversity compared to controls.
• Elevated levels of propionic acid and a specific butyric acid/valerate difference were observed in IBS patients.
• A machine learning model based on key bacterial features achieved >80% sensitivity and >90% specificity for IBS prediction.

Conclusions:

• Gut microbiome analysis combined with machine learning provides a robust method for objective IBS identification.
• This approach can complement or improve current diagnostic methods for IBS.
• Microbial and metabolic profiles offer potential biomarkers for IBS.