Answers are grounded in this paper and cite their sources.
Background Esophageal carcinoma (EC) is the seventh leading cause of cancer-related mortality worldwide, yet its immune-regulatory circuitry remains poorly defined. We aimed to prioritise immune-cell-expressed candidate genes for EC by integrating whole-blood expression quantitative trait loci (eQTL) summary statistics with Bayesian colocalisation, two-sample Mendelian randomisation (MR), and single-cell eGene annotation from OneK1K. Methods Summary-data-based Mendelian randomisation (SMR) used eQTLGen whole-blood cis-eQTL and EC GWAS summary statistics (GCST90018841; 998 European-ancestry and 1,388 East-Asian cases; 475,308 European-ancestry and 159,201 East-Asian controls), with 14 OneK1K immune-cell-subset eGene lists (26,597 entries) defining testing scopes. Benjamini-Hochberg correction covered all tests. Colocalisation used coloc v5.2.3, all regional SNPs and twelve prior settings; two-sample MR was repeated after standard LD clumping (r 2 < 0.001, 10,000 kb; 1000 Genomes EUR). Because the outcome GWAS is shared, MR is an internal consistency and directionality check, not independent replication. Candidates were tiered under pre-specified criteria and annotated by STRING, DGIdb, TRRUST, GeneMANIA, CTD and UniProt. Differential expression was assessed in GSE20347 (Affymetrix microarray; 17 paired ESCC/normal; patient-matched limma) and TCGA-ESCA (RNA-seq; DESeq2). Results Across 5,331 unique genes (14,491 gene-by-subset tests), 571 gene-by-subset associations (238 unique genes) were nominally significant, and none survived global FDR correction (smallest q = 0.08); the study is therefore framed as exploratory. Of the 22 candidates carried forward from the previous sentinel-SNP screen, multi-SNP colocalisation supported seven at PPH4 > 0.5 under default priors; PPH4 ranged from 0.06 to 0.99 across prior settings and no gene maintained PPH4 > 0.8 in all settings. Six candidates with significant MR after standard clumping formed the primary panel: NFAM1 (OR = 2.32) and SRI (1.68) were risk-increasing, whereas ZC3H7A (0.63), UBE2Q2 (0.72), TSPAN32 (0.85) and ASCL2 (0.86) were protective. Transcriptomic support was partial and gene-specific (ASCL2 and SRI up-regulated in TCGA-ESCA; TSPAN32 down-regulated in both cohorts; ZC3H7A elevated in GSE20347 but decreased in TCGA-ESCA). One medium-confidence STRING interaction (TSPAN32–ASCL2, score 0.65) and four approved drugs targeting SRI were identified. Conclusion This exploratory, sensitivity-tested analysis nominates six immune-cell-expressed candidate genes for EC and provides the complete, auditable evidence trail; experimental validation is required to establish any functional or clinical role.
No takes yet. Share an insight, caveat, or question.
Liu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: