Related Experiment Video
Updated: Jun 25, 2025
Analysis of Fecal Microbiota Dynamics in Lupus-Prone Mice Using a Simple, Cost-Effective DNA Isolation Method
Published on: May 2, 2022
Machine learning-based characterization of the gut microbiome associated with the progression of primary biliary
Qi Wang1, Xiaomeng Tang1, Wenying Qiao2
1Beijing Key Laboratory of Emerging Infectious Diseases, Institute of Infectious Diseases, Beijing Ditan Hospital, Capital Medical University, Beijing, PR China; Beijing Institute of Infectious Diseases, Beijing, PR China; National Center for Infectious Diseases, Beijing Ditan Hospital, Capital Medical University, Beijing, PR China.
Background:
Primary biliary cholangitis (PBC) is associated closely with the gut microbiota. This study aimed to explore the characteristics of the gut microbiota after the progress of PBC to cirrhosis.
Method:
This study focuses on utilizing the 16S rRNA gene sequencing method to screen for differences in gut microbiota in PBC patients who progress to cirrhosis. Then, we divided the data into training and verification sets and used seven different machine learning (ML) models to validate them respectively, calculating and comparing the accuracy, F1 score, precision, and recall, and screening the dominant intestinal flora affecting PBC cirrhosis.
Result:
PBC cirrhosis patients showed decreased diversity and richness of gut microbiota. Additionally, there are alterations in the composition of gut microbiota in PBC cirrhosis patients. The abundance of Faecalibacterium and Gemmiger bacteria significantly decreases, while the abundance of Veillonella and Streptococcus significantly increases. Furthermore, machine learning methods identify Streptococcus and Gemmiger as the predominant gut microbiota in PBC patients with cirrhosis, serving as non-invasive biomarkers (AUC = 0.902).
Conclusion:
Our study revealed that PBC cirrhosis patients gut microbiota composition and function have significantly changed. Streptococcus and Gemmiger may become a non-invasive biomarker for predicting the progression of PBC progress to cirrhosis.
