Multi-omic and machine learning analysis of mitochondrial RNA modification genes in lung adenocarcinoma for prognostic and therapeutic implications

Abstract

Lung cancer remains the leading cause of cancer-related deaths, driven by complex pathogenesis and poor prognosis. Recognizing the pivotal role of mitochondrial RNA modifications (MRM) in cancer progression, this study aims to provide a comprehensive analysis of MRM-related genes and their clinical relevance in lung adenocarcinoma (LUAD). Integrating multi-omic datasets, we systematically explored the molecular features of MRM-related genes across various cancers and identified distinct expression patterns and prognostic associations. Single-cell analysis further reveals MRM-driven cell-cell interactions and pathway activation, particularly in cycling and epithelial cells. Using advanced machine learning techniques, we developed a novel prognostic signature—the Mitochondrial RNA Modification-related Signature (MRMS)—comprising nine genes: TXN, LDHA, HMGA1, SFTPB, KRT8, ALG3, S100A16, HSPD1, and ALDOA. The MRMS demonstrates superior predictive performance for LUAD survival compared to previously reported models. Our findings uniquely link MRMS to increased tumor mutational burden, genetic instability, and an immunosuppressive tumor microenvironment characterized by reduced immune cell infiltration and elevated tumor purity. Additionally, MRMS is associated with immunotherapy-related features, suggesting its potential in predicting treatment response. Experimental validation identified ALG3 as an oncogenic driver in LUAD, influencing tumor cell proliferation, migration, and invasion. In conclusion, this study establishes MRMS as a robust prognostic biomarker and highlights its dual role in shaping the tumor immune microenvironment and guiding therapeutic strategies. These findings provide novel insights into mitochondrial RNA modifications and their potential applications in personalized treatment for LUAD.