To identify noninvasive biomarkers of ulcerative colitis (UC) severity and disease extent using serum metabolic profiling and machine learning.
Key Findings:
220 differential metabolites were identified between UC patients and healthy controls.
8 essential metabolites were screened to distinguish UC patients from healthy controls.
The random forest model achieved 100% prediction accuracy across all training sets.
Tridecanoic acid levels were lower and pelargonic acid levels were higher in patients with extensive colitis.
Asparaginyl valine levels were significantly lower in patients with rectal UC compared to other groups.
23, 6, and 6 differential metabolites were identified between groups E1 and E2, E1 and E3, and E2 and E3, respectively.
Interpretation:
The study reveals a distinct metabolic profile associated with different extents of UC, highlighting the potential of metabolites as biomarkers for disease severity and progression.
Limitations:
The study's sample size and demographic diversity may limit the generalizability of findings, particularly across different populations.
Complex data processing and interpretation in metabolomics can pose challenges, potentially affecting the reliability of results.
Conclusion:
This research underscores the utility of combining metabolomics with machine learning to identify biomarkers for UC severity and disease extent, which may enhance therapeutic decision-making and risk stratification.
