Machine learning-based algorithm identifies key mitochondria-related genes in non-alcoholic steatohepatitis

Area of Science:

• Hepatology
• Mitochondrial Biology
• Genetics

Background:

• Hepatocyte mitochondrial dysfunction is implicated in non-alcoholic steatohepatitis (NASH) pathogenesis.
• This dysfunction contributes to abnormal lipid metabolism, redox imbalance, and cell death.
• Identifying key mitochondrial genes in NASH may reveal novel therapeutic targets.

Purpose of the Study:

• To identify critical mitochondrial hub genes associated with non-alcoholic steatohepatitis (NASH).
• To explore the role of these genes in NASH progression and disease severity.

Main Methods:

• A comprehensive analysis of 134 algorithms was performed to identify mitochondrial hub genes in NASH.
• The Random Forest (RF) algorithm demonstrated the highest efficacy in this identification process.

Main Results:

• The RF algorithm identified three upregulated genes: Aldo-keto reductase family 1 member B10 (AKR1B10), thymidylate synthase (TYMS), and triggering receptor expressed in myeloid cell 2 (TREM2).
• These genes positively correlate with inflammation, lipid synthesis, fibrosis, and NASH activity scores.
• The identified genes accurately stratified NASH patients into distinct clusters based on disease severity.

Conclusions:

• AKR1B10, TYMS, and TREM2 are pivotal mitochondrial genes involved in NASH progression.
• These genes serve as potential biomarkers for disease severity and therapeutic targets.