Identification of palmitoylation-related diagnostic biomarkers in dilated cardiomyopathy through multi-omics integration and experimental validation

Dilated cardiomyopathy (DCM) is a progressive myocardial disorder characterized by ventricular dilation and systolic dysfunction, with limited options for early molecular diagnosis. Protein palmitoylation is a reversible lipid post-translational modification involved in membrane localization, protein stability, and signal transduction; however, direct palmitoylation profiling in DCM remains limited. Here, we performed an integrative transcriptomic analysis combining bulk and single-nucleus RNA sequencing datasets to identify palmitoylation-related gene (PRG) transcriptional signatures associated with DCM. Through differential expression analysis, functional enrichment, consensus clustering, and machine-learning-based feature selection, five core PRGs (GOLGA7, LYPLA1, ZDHHC3, ZDHHC14, and ZDHHC15) were identified and used to construct a diagnostic model, which was evaluated across five independent external cohorts. Immune infiltration analysis and PRG-based clustering suggested two transcriptional groups with different immune- and metabolism-related features. Single-nucleus analysis suggested cell type-specific differences in PRG-associated transcriptional scores and core gene expression patterns, particularly in fibroblasts and cardiomyocytes. Drug–gene interaction analysis and molecular docking prioritized valproic acid as an exploratory candidate compound. The mRNA expression patterns of core genes were further validated in human myocardial tissues and a doxorubicin-induced mouse DCM model. Overall, this study characterizes PRG-associated transcriptional signatures in DCM and provides a hypothesis-generating framework for future mechanistic validation.