scMRdb: a comprehensive resource for cell-type-resolved Mendelian randomization of complex traits.

scMRdb harmonizes 5206 European and East Asian GWAS datasets with 585 sc-eQTL datasets across 11 tissues and cell systems, enabling cell-type-resolved MR across 110 cell types.

• Why it matters: Bulk-tissue eQTLs can dilute cell-type-specific transcriptional effects, and sc-eQTL MR evidence is fragmented across studies, limiting standardized cross–cell-type comparisons.
• What they did: The resource applies a uniform pipeline to evaluate gene–trait pairs via MR with Steiger directionality testing, then uses colocalization and six confidence tiers plus drug-target annotation.
• The result: It systematically recovers candidate causal genes supported by cell-type-resolved regulatory evidence that is often diluted or obscured in bulk-tissue analyses, supporting mechanistic hypothesis generation and therapeutic target prioritization.