Characterization of m6A ‐Related Genes in Tumor‐Associated Macrophages for Prognosis, Immunotherapy, and Drug Prediction in Lung Adenocarcinomas Based on Machine Learning Algorithms

Tumor-associated macrophages (TAMs) are a vital immune component within the tumor microenvironment (TME) of lung adenocarcinoma (LUAD), exerting significant influence on tumor growth, metastasis, and drug resistance. N6-methyladenosine (m6A) modifications regulate gene expression and various facets of cancer biology; nonetheless, the mechanisms by which they modulate gene expression in TAMs and their impact on LUAD progression remain inadequately elucidated. Single-cell transcriptome analysis identified the macrophage m6A-related genes (MMRGs) with high expression in TAMs and linked to m6A modifications in LUAD. The MMRGs were employed to construct 801 prognostic models by 13 different machine learning (ML) algorithms. An integrative multi-omics approach was utilized to analyze the potential biological functions of MMRGs. Five ML algorithms were employed to discover potential biomarkers for patient stratification and precision therapy. Potential drugs for treating LUAD were identified and assessed by molecular docking and molecular dynamics, and ML was employed to determine the most promising candidates. Immunohistochemistry and immunofluorescence staining were conducted to assess MMRG expression in LUAD tissues. Seventeen MMRGs were identified in LUAD and subsequently employed to construct a prognostic model for patient stratification into high-risk and low-risk groups. The impact of MMRG expression on various tumor immune phenotypes, such as tumor stemness, heterogeneity, hallmark pathway enrichment, TAM infiltration, and immune landscape, was thoroughly characterized. PIM3, HMGB2, DUSP2, NR4A2, and others were recognized as promising biomarkers for patient classification and precision therapy. Furthermore, it was predicted that drugs such as BRD9876 and MK-1775 would demonstrate therapeutic efficacy in treating LUAD, and drugs showing potential binding with DUSP2, ZNF331, FLT, and LYZ were identified. Finally, experimental validation was conducted to confirm the protein expression and distribution of DUSP2 and NR4A2 in tissues of LUAD. Our study offers valuable insights into the biological significance of MMRGs, shedding light on novel mechanisms of tumor development and immune evasion in LUAD. Furthermore, our findings have identified potential biomarkers, drug candidates, and therapeutic targets that may improve the management of LUAD in the future.