基于多组学孟德尔随机化的胃癌关键血浆蛋白靶标发现与药物预测

Identification of key plasma protein targets and drug prediction for gastric cancer based on multi-omics Mendelian randomization

摘要:

背景与目的 胃癌是消化系统常见的高致死性恶性肿瘤，现有治疗手段疗效有限，亟须发现新的治疗靶点。本研究基于孟德尔随机化方法，整合全基因组关联研究（GWAS）与血浆蛋白定量性状位点（pQTL）数据，系统筛选并验证与胃癌发生存在因果关联的血浆蛋白，为靶向治疗提供理论依据。方法 采用双样本孟德尔随机化方法，结合胃癌GWAS与血浆pQTL数据开展因果推断，并引入外部独立数据集进行验证。通过多维度敏感性分析（包括反向因果检验、贝叶斯共定位分析及表型扫描）评估结果稳健性。基于STRING数据库构建蛋白互作网络，解析候选蛋白的通路机制，并利用DrugBank数据库预测潜在靶向药物。结果 初步识别出16种与胃癌风险明显相关的因果血浆蛋白。经外部验证和敏感性分析，ICAM2、IGF1R、LIFR和MET被确认为关键候选靶标。药物数据库提示，靶向IGF1R的dalotuzumab及可能调控ICAM2通路的efalizumab等药物具有潜在治疗价值。结论 本研究通过多组学整合的孟德尔随机化分析，系统鉴定了4种与胃癌具有稳定因果关系的血浆蛋白（ICAM2、IGF1R、LIFR和MET）。上述靶点不仅为揭示胃癌的分子病理机制提供了新视角，也为后续靶向药物的研发与个体化治疗策略的制定提供了理论支撑。

Abstract:

Background and Aims Gastric cancer is a common and highly lethal malignancy of the digestive system. The efficacy of current treatment strategies remains limited, highlighting the urgent need to identify novel therapeutic targets. This study employed a Mendelian randomization (MR) approach to integrate GWAS data with pQTL data, aiming to systematically identify and validate plasma proteins that are causally associated with gastric cancer, thereby providing a theoretical basis for targeted therapy.Methods A two-sample Mendelian randomization analysis was conducted using GWAS data on gastric cancer and plasma pQTL datasets to infer causal relationships. External independent datasets were used for validation. Multi-dimensional sensitivity analyses-including reverse causality testing, Bayesian colocalization, and phenome-wide scans-were performed to ensure the robustness of the findings. Protein-protein interaction networks were constructed via the STRING database to elucidate the biological pathways of candidate proteins, and the DrugBank database was utilized to predict potential therapeutic agents.Results A total of 16 plasma proteins were initially identified as causally associated with the risk of gastric cancer. After external validation and sensitivity analyses, ICAM2, IGF1R, LIFR, and MET were confirmed as key candidate targets. Drug database analysis indicated that dalotuzumab (targeting IGF1R) and efalizumab (potentially modulating the ICAM2 pathway) may have therapeutic potential.Conclusion Through a multi-omics Mendelian randomization framework, this study systematically identified four plasma proteins-ICAM2, IGF1R, LIFR, and MET-that exhibit stable causal associations with gastric cancer. These targets offer novel insights into the molecular pathogenesis of gastric cancer and provide a theoretical foundation for developing targeted drugs and personalized treatment strategies.