Dermatomyositis (DM) and polymyositis (PM) are rare idiopathic inflammatory myopathies (IIMs) characterized by systemic inflammation and muscle weakness. While genome-wide association studies (GWAS) have identified numerous susceptibility loci, the biological translation of these non-coding variants into cell-type-specific mechanisms remains a significant challenge. This study aims to bridge this gap by identifying causal genes and therapeutic targets at single-cell resolution.

We integrated single-cell expression quantitative trait loci (sc-eQTL) data from 1.27 million PBMCs (OneK1K cohort) with DM/PM GWAS summary statistics (FinnGen R12) using a single-cell transcriptome-wide association study (scTWAS) framework. Causal relationships were inferred through two-sample Mendelian randomization (MR) and validated using an independent sc-eQTL dataset (1M-scBloodNL). High-confidence causal drivers were prioritized via Bayesian colocalization analysis (PP.H4 > 0.8). Furthermore, we constructed protein-protein interaction (PPI) networks, conducted Phenome-wide association studies (PheWAS) to assess safety profiles, and performed drug repurposing analysis using the DSigDB database.

MR analysis identified 204 unique eGenes with significant causal associations (FDR < 0.05) across 14 immune cell types, highlighting the involvement of both innate (monocytes, NK cells) and adaptive (B cells, T cell subsets) immunity. Colocalization analysis prioritized BTN3A2, HLA-DRB1, and HLA-C as robust, cell-type-specific drivers. BTN3A2 exhibited consistent protective effects across T and B cell lineages, whereas HLA-DRB1 acted as a stable risk factor in monocytes and immature B cells. Notably, HLA-C was identified as a terminal differentiation-dependent risk factor specifically in plasma cells (PP.H4 = 0.91), suggesting a novel pathogenic role for the B-cell-to-plasma-cell transition. Drug enrichment analysis identified tofacitinib (Padj = 9.22 × 10⁻³¹) and LY-294002 as top therapeutic candidates. PheWAS confirmed that targeting these prioritized genes presents minimal risk of non-immune off-target effects.

This study provides the first cell-type-resolved genetic map of DM/PM, uncovering lineage-specific regulatory mechanisms. The identification of BTN3A2 as an immune stabilizer and the plasma-cell-specific role of HLA-C offer precise avenues for immunomodulation. These findings validate existing treatments like JAK inhibitors and propose novel strategies, such as plasma-cell-targeted therapies (e.g., BCMA-targeted agents), to achieve deeper remission in refractory patients.