Application of large-scale and multicohort plasma proteomics data to discover novel causal proteins in gastric cancer

Abstract

PurposesGastric cancer (GC) is one of the most common malignant tumors threatening human beings and has a poor prognosis. Therefore, exploring unveiled biomarkers or therapeutic targets for the diagnosis and treatment of GC is crucial.MethodsA total of 5772 protein quantitative trait loci (pQTL) were aggregated from four latest large-scale proteomics cohorts. Two-sample Mendelian randomization (two-sample MR) was utilized to identify the causal effect of blood plasma proteins on GC. Heterogeneity, pleiotropy, and directionality analyses were employed to evaluate proteins identified via two-sample MR. The robustness of results was further validated via colocalization. The drug targets of proteins were evaluated to reveal the compounds that can interfere with these proteins.ResultsTen proteins with potential causations in relation to GC were identified: LY6D, SLURP1, MLN, PSCA, THSD1, CFTR, PPM1B, KDM3A, TSC1, and HCG22. Among these proteins, LY6D, SLURP1, and THSD1 were considered as the most reliable biomarkers of GC due to their significant H4 posterior probabilities in colocalization analysis and absence of pleiotropy. Compound 35, nitrosamide, and 0175029-0000 were potential drugs or small molecules targeting LY6D, SLURP1, and THSD1, respectively.ConclusionThis study identified several plasma proteins as potential biomarkers of GC and provided data support and new insights into the early diagnosis, intervention, and therapeutic targets of GC.