Refined expression quantitative trait locus analysis on adenocarcinoma at the gastroesophageal junction reveals susceptibility and prognostic markers.

This study aimed to explore cell type level expression quantitative trait loci (eQTL) in adenocarcinoma at the gastroesophageal junction (ACGEJ) and identify susceptibility and prognosis markers. Whole-genome sequencing (WGS) was performed on 120 paired samples from Chinese ACGEJ patients. Germline mutations were detected by GATK tools. RNA sequencing (RNA-seq) data on ACGEJ samples were taken from our previous studies. Public single-cell RNA sequencing (scRNA-seq) data were used to produce the proportion of epithelial cells. Matrix eQTL and a linear mixed model were used to identify condition-specific cis-eQTLs. The R package coloc was used to perform co-localization analysis with the public data of genome-wide association studies (GWASs). Log-rank and Cox regression tests were used to identify survival-associated eQTL and genes. Functions of candidate risk loci were explored by experimental validation. Refined eQTL analyses of paired ACGEJ samples were performed and 2,036 potential ACGEJ-specific eQTLs with East Asian specificity were identified in total. ACGEJ-gain eQTLs were enriched at promoter regions more than ACGEJ-loss eQTLs. rs658524 was identified as the top eQTL close to the transcription start site of its paired gene (). rs2240191-, rs4236599-, rs4947311-, rs13134812-, and rs17508585- were identified as ACGEJ-specific susceptibility eQTLs. rs309483- was associated with the overall survival of ACGEJ patients. We explored functions of candidate eQTLs such as rs658524, rs309483, rs2240191, and rs4947311 by experimental validation. This study provides new risk loci for ACGEJ susceptibility and effective disease prognosis biomarkers.