What question did this study set out to answer?

The research aims to identify mitophagy-related core genes and uncover their roles in non-alcoholic fatty liver disease (NAFLD).

March 26, 2026Journal of Inflammation Research0 citationsOpen Access

Integrated Machine Learning and Multi-Omics Analysis Identifies Mitophagy-Related Core Genes and Mechanisms in Non-Alcoholic Fatty Liver Disease

YYYu Yuan TZTianyu Zhang CSChao Song

Key Points

The research aims to identify mitophagy-related core genes and uncover their roles in non-alcoholic fatty liver disease (NAFLD).
Integrated multiple NAFLD transcriptomic datasets and single-cell RNA sequencing data.
Employed machine learning algorithms and bioinformatics for core gene identification.
Conducted immune infiltration analysis, PPI network construction, and cell-cell communication analysis.
Validated findings with in vitro experiments such as Western Blot and qRT-PCR.
Identified five key genes: IGF1, MYH11, HYOU1, SPATA18, and SCD, with a training set AUC of 0.974.
These genes are enriched in processes like mitophagy, lipid metabolism, and endoplasmic reticulum stress.
Demonstrated roles in altering the NAFLD immune microenvironment, increasing macrophage M2 and T cell infiltration.
Revealed gene expression heterogeneity among different cell types in the disease context.

Abstract

Objective: This study aimed to systematically screen for mitophagy-related core genes in Non-alcoholic fatty liver disease (NAFLD), elucidate their specific molecular regulatory network, and investigate their functional mechanisms and roles within the immune microenvironment to provide novel targets for disease diagnosis and therapy. Methods: Multiple NAFLD transcriptomic datasets and single-cell RNA sequencing data from the GEO database were integrated. Bioinformatics analysis, Weighted Gene Co-expression Network Analysis (WGCNA), and 11 machine learning algorithms were employed for core gene screening. Functional mechanisms and immune microenvironment characteristics were further investigated using SHAP model interpretability analysis, including detailed immune infiltration analysis, PPI network construction, GSEA, single-cell trajectory inference, and cell-cell communication analysis. Reverse network pharmacology and molecular docking predicted potential targeted compounds. In vitro experiments (Western Blot, qRT-PCR, JC-1 staining) validated core gene expression and mitophagy levels. Results: Five key genes-IGF1, MYH11, HYOU1, SPATA18, and SCD-were identified, demonstrating excellent disease discrimination across multiple cohorts (training set AUC=0.974). These genes were significantly enriched in processes like endoplasmic reticulum stress, mitophagy, and lipid metabolism. Critically, they played crucial roles in reshaping the NAFLD immune microenvironment, characterized by increased macrophage M2 polarization and T cell infiltration, linking mitochondrial dysfunction to inflammatory response. Single-cell analysis revealed their expression heterogeneity across hepatocytes, macrophages, and T cells, along with their involvement in intercellular communication patterns. Experimental validation confirmed aberrant core gene expression and altered mitophagy levels in NAFLD cell models. Conclusion: This study systematically delineates the regulatory network of mitophagy-related core genes in NAFLD and the resultant inflammatory immune microenvironment, offering novel insights and data support for elucidating disease mechanisms, developing early diagnostic biomarkers, and formulating precise therapeutic strategies. Keywords: non-alcoholic fatty liver disease, mitophagy, machine learning, bioinformatics

Perguntar à IA

Bookmark

View Full Paper