Identification of novel high-risk genes in gastric cancer through single-cell RNA sequencing, eQTL Mendelian randomization, and in vitro validation.

[BACKGROUND] Current targeted therapies for gastric cancer have limited efficacy, and recently discovered markers have not significantly improved survival rates in patients with gastric cancer. Therefore, it is imperative to identify more specific genes associated with the occurrence and progression of gastric cancer to achieve prevention and treatment. The aim of this study is to discover high-risk genes for gastric cancer by integrating single-cell transcriptomics and Mendelian randomization (MR) analysis.

[METHODS] This study integrates gastric cancer genome-wide association study (GWAS) data, single-cell transcriptomics (sc-RNA-seq), and expression quantitative trait loci (eQTL) data for analysis, and employs two-sample MR to elucidate the causal relationships between genes and gastric cancer, thereby identifying high-risk genes for gastric cancer. Subsequently, in vitro cellular experiments are conducted to validate the transcriptional expression levels of these genes.

[RESULTS] After quality control of the sc-RNA-seq data, we identified 2463 markers for gastric cancer cell subtypes. Subsequently, we utilized eQTL data and GWAS data for gastric cancer to perform MR analysis, yielding 149 genes with a causal relationship with gastric cancer. By applying log2FC filtering, we ultimately identified 5 high-risk gastric cancer genes: SORBS3, RMND5A, FBXO6, LPGAT1, and EPHB4. Finally, in vitro validation confirmed the differential expression of these 5 high-risk genes between normal gastric epithelial cell lines and gastric cancer cell lines.

[CONCLUSIONS] Our study reveals previously unattended high-risk gastric cancer genes, potentially offering new directions and evidence for the molecular diagnosis and treatment of gastric cancer.