INTEGRATIVE TRANSCRIPTOMIC AND MACHINE LEARNING ANALYSIS REVEALS CANDIDATE HUB GENES IN DUCHENNE MUSCULAR DYSTROPHY

Duchenne muscular dystrophy (DMD) is a fatal X-linked disorder characterized by progressive muscle degeneration and cardiomyopathy, for which reliable early biomarkers remain limited. This study aimed to identify key molecular biomarkers and regulatory networks in DMD cardiac fibroblasts using integrative transcriptomic and machine learning approaches. RNA-sequencing data (GSE237014) from induced pluripotent stem cell–derived cardiac fibroblasts of four DMD patients and four healthy controls were analyzed to identify differentially expressed genes (DEGs). Protein–protein interaction networks were constructed using STRING, and hub genes were identified via CytoHubba. Machine learning models, including Gradient Boosting Classifier, Random Forest, and Ridge Classifier, were applied to prioritize candidate biomarkers, with model interpretability assessed using Local Interpretable Model-Agnostic Explanations (LIME). Regulatory networks integrating transcription factors (TFs) and microRNAs (miRNAs) were constructed to elucidate upstream regulatory mechanisms. A total of 1,520 DEGs were identified, comprising 652 upregulated and 868 downregulated genes. Fifty hub genes were significantly enriched in mitotic regulation, spindle assembly, and cell cycle–related pathways. Integrated machine learning analyses consistently highlighted five hub genes—CCNB1, CDC20, CEP55, HNRNPA1, and HNRNPM—as key discriminators between DMD and healthy controls, based on convergent LIME and Z-score-based feature-importance analyses. Given the limited sample size (n = 8), these findings should be interpreted as exploratory candidate biomarkers requiring validation in larger, independent cohorts. Regulatory network analysis identified TFs MYC, NFYA, and YBX1, along with miRNAs miR-15b-5p, miR-16-5p, and miR-92a-3p, as central upstream regulators. This integrative framework identifies reproducible candidate hub genes and regulatory circuits, underscoring mitotic dysregulation as a hallmark of DMD cardiac fibroblasts and nominating potential biomarkers and therapeutic targets.