<p>Mitochondrial dysfunction is a central pathological feature in the comorbidity of Type 2 Diabetes Mellitus (T2DM) and Non-Alcoholic Fatty Liver Disease (NAFLD). This study employed an integrative genomics approach to identify key mitochondrial metabolism-related genes (MMRGs), including genes annotated in mitochondrial pathways) serving as shared diagnostic and therapeutic targets. By integrating transcriptomic data from human NAFLD and T2DM datasets and applying machine learning algorithms with cross-dataset validation, we identified three pivotal biomarkers: <i>CYP4F2</i>,<i> CYP8B1</i>, and <i>SP1</i>. Nomogram models incorporating these genes demonstrated high predictive accuracy for both conditions. Functional enrichment analyses linked these genomic markers to critical pathways, including actin cytoskeleton regulation in NAFLD and phagosome signaling in T2DM. Immune profiling further revealed significant associations, including CYP8B1 correlating with monocytes in NAFLD and neutrophils in T2DM. Computational drug prediction identified potential therapeutic candidates targeting each biomarker. Crucially, the diagnostic utility of these markers was translated from transcript to protein via ELISA validation, which confirmed significant dysregulation in serum levels that correlated with NAFLD severity in diabetic patients. In conclusion, this study establishes CYP4F2, CYP8B1, and SP1 as pivotal genomic bridges in metabolism-related pathways linking T2DM and NAFLD, providing a foundation for novel diagnostic strategies and targeted precision therapies.</p> Graphical Abstract <p></p>