Abstract
Abstract
Cutaneous melanoma is characterized by marked cellular heterogeneity and therapeutic resistance, complicating the interpretation of tissue-level transcriptomic signals. Bulk RNA sequencing provides cohort-scale detection of melanoma-associated transcriptional programs but lacks cellular resolution, whereas single-cell RNA sequencing enables cell type-specific characterization in smaller cohorts. Here, we used a concordance-based bulk-informed single-cell transcriptomic framework to prioritize cell type-resolved melanoma-associated genes and candidate therapeutic vulnerabilities. Differential expression results from multiple independent bulk RNA-seq cohorts were integrated with cell type-specific single-cell profiles by retaining genes with concordant regulatory direction across data types. Melanoma-derived melanocytes showed prominent enrichment of cell-cycle, mitotic checkpoint, and DNA-replication programs, whereas immune and stromal populations contributed primarily antigen-processing and interferon-related signals. In macrophages, concordant upregulation of LGALS9 and NFKBIE highlighted a potential immunoregulatory program consistent with attenuation of T-cell activity. Integration with DepMap dependency and protein-network centrality analyses further prioritized CCNB1, MAD2L1, MCM3, MCM6, and PCNA as candidate melanoma-associated vulnerabilities. Drug-signature analysis further identified FDA-approved compounds predicted to reverse the prioritized five-gene melanoma-associated transcriptional signature. These findings provide a cell type-resolved prioritization of melanoma-associated transcriptional programs and generate candidate biomarkers and therapeutic hypotheses for further clinical and experimental validation.