Integrated multi-omics analysis identifies key biomarkers associated with post-translational modifications and RNA methylation in clear cell renal cell carcinoma.

backgroundThis study aimed to identify clinically relevant molecular signatures and biomarkers associated with post-translational modifications (PTMs) and RNA methylation in clear cell renal cell carcinoma (ccRCC) by integrating multi-omics data to elucidate tumorigenesis mechanisms and tumor microenvironment dynamics for potential diagnostic and therapeutic advancements.

methodsWe analyzed bulk RNA-sequencing data from five GEO datasets, GSE16449, GSE46699, GSE53000, GSE53757, and GSE66272, with batch-effect correction using the sva package and single-cell RNA-seq data processed via Seurat v4 with Harmony integration. Differential expression analysis using limma identified PTM- and methylation-related gene signatures. Functional enrichment using clusterProfiler and Weighted Gene Co-expression Network Analysis (WGCNA) revealed key modules linked to 20 PTM types and four RNA methylation patterns, m1A, m5C, m6A, and m7G. Machine learning using LASSO, SVM, and Random Forest, along with SHAP-based random forest modeling, selected and evaluated biomarkers. Immune infiltration was assessed via ssGSEA, and consensus clustering defined molecular subtypes. Statistical analyses using Wilcoxon and Kruskal-Wallis tests with FDR correction ensured robustness.

resultsWe identified 2,779 differentially expressed genes, including 14 significant PTM and methylation signatures including 11 PTMs, 3 methylation types, enriched in PI3K-Akt signaling and immune response pathways. WGCNA revealed four disease-associated modules tied to PTMs and RNA methylation. Single-cell analysis delineated 16 cell types, with T cells dominant in tumors and enhanced cell-cell interactions in high-modification groups. Machine learning identified PDIA3, STT3A, and USP4 as core biomarkers, with SHAP confirming STT3A's predictive strength. Biomarkers showed elevated expression in ccRCC, correlating with dendritic and T cell infiltration. Consensus clustering defined two subtypes: C2 exhibited higher PTM/methylation-related gene expression, oncogenic pathway enrichment, and lower immune infiltration compared to C1.

conclusionThis integrative multi-omics framework identifies PDIA3, STT3A, and USP4 as key biomarkers linked to PTMs and RNA methylation, delineating two molecular subtypes. These findings enhance understanding of ccRCC's molecular and immune landscape, offering insights for improved diagnostic and therapeutic strategie.