GEOapplication/xmlftp://ftp.ncbi.nlm.nih.gov/geo/series/GSE327nnn/GSE327937/primaryOK200TranscriptomicsHomo sapiensExpression profiling by high throughput sequencinghttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE327937GEOGSEfalseMiT Fusions, TSC1–TSC2 Divergence, and Stem-like Programs Reveal Distinct Origins and Vulnerabilities in PEComa [scRNA-seq]Perivascular epithelioid cell neoplasms (PEComas) are ultra-rare mesenchymal tumors lacking a molecular classification to guide therapy. Here we perform comprehensive multi-omic profiling of an unselected PEComa cohort. We identify frequent MITF rearrangements involving actin gene partners (ACTA2, ACTG1 and ACTB). Anatomical stratification reveals cyclin-dependent kinase module mutations in gynecologic tumors, whereas soft tissue, gastrointestinal, and pelvic tumors lacked mTOR pathway alterations but are enriched for TFE3/MITF rearrangements. Transcriptomic analysis defines four subtypes with distinct lineage programs—melanocytic, mesenchymal, or adipogenic—as well as unique mutational patterns and clinical behaviors. Notably, an aggressive stem-like subtype enriched for TP53/RB1 mutations exhibits high proliferation, activation of embryonic and Hedgehog signaling, immune infiltration, and resistance to mTOR inhibitors, but potential responsiveness to immunotherapy. Single-nucleus RNA sequencing reveals intra-tumoral heterogeneity within this subtype, including divergent inflammatory states. Together, these findings establish a molecular classification framework and identify actionable vulnerabilities in PEComa.2026/04/18GSE327937GSM9669039GSM9669042GSM9669041GSM9669043GSM966904024676327937Homo sapiens