Data Sheet 1_Analysis of shared pathogenic mechanisms and drug targets in myocardial infarction and gastric cancer based on transcriptomics and machine learning.csv

BackgroundRecent studies have suggested a potential association between gastric cancer (GC) and myocardial infarction (MI), with shared pathogenic factors. This study aimed to identify these common factors and potential pharmacologic targets.
MethodsData from the IEU Open GWAS project were used. Two-sample Mendelian randomization (MR) analysis was used to explore the causal link between MI and GC. Transcriptome analysis identified common differentially expressed genes, followed by enrichment analysis. Drug target MR analysis and eQTLs validated these associations with GC, and the Steiger direction test confirmed their direction. The random forest and Lasso algorithms were used to identify genes with diagnostic value, leading to nomogram construction. The performance of the model was evaluated via ROC, calibration, and decision curves. Correlations between diagnostic genes and immune cell infiltration were analyzed.
ResultsMI was linked to increased GC risk (OR=1.112, P=0.04). Seventy-four genes, which are related mainly to ubiquitin-dependent proteasome pathways, were commonly differentially expressed between MI and GC. Nine genes were consistently associated with GC, and eight had diagnostic value. The nomogram built on these eight genes had strong predictive performance (AUC=0.950, validation set AUC=0.957). Immune cell infiltration analysis revealed significant correlations between several genes and immune cells, such as T cells, macrophages, neutrophils, B cells, and dendritic cells.
ConclusionMI is associated with an increased risk of developing GC, and both share common pathogenic factors. The nomogram constructed based on 8 genes with diagnostic value had good predictive performance.

背景：近期研究提示胃癌（gastric cancer, GC）与心肌梗死（myocardial infarction, MI）之间存在潜在关联，且二者共享致病因素。本研究旨在明确这些共同致病因素及潜在药理学靶点。
方法：本研究使用了IEU开放全基因组关联研究（Genome-Wide Association Study, GWAS）项目的数据。采用双样本孟德尔随机化（Mendelian randomization, MR）分析探究MI与GC之间的因果关联。通过转录组分析筛选二者共有的差异表达基因，随后进行富集分析。借助药物靶点MR分析与表达数量性状基因座（expression quantitative trait locus, eQTL）验证上述基因与GC的关联，并通过Steiger方向检验确认关联方向。采用随机森林与Lasso算法筛选具有诊断价值的基因，进而构建列线图（nomogram）。通过ROC曲线（receiver operating characteristic curve）、校准曲线与决策曲线评估模型性能。分析诊断基因与免疫细胞浸润之间的相关性。
结果：MI与GC发病风险升高显著相关（比值比OR=1.112，P=0.04）。在MI与GC之间共筛选出74个差异表达基因，这些基因主要富集于泛素依赖性蛋白酶体通路。其中9个基因与GC持续相关，8个基因具有诊断价值。基于这8个基因构建的列线图表现出优异的预测性能（训练集曲线下面积AUC=0.950，验证集AUC=0.957）。免疫细胞浸润分析显示，多个基因与T细胞、巨噬细胞、中性粒细胞、B细胞及树突状细胞等免疫细胞存在显著相关性。
结论：MI与GC发病风险升高存在关联，且二者共享共同致病因素。基于8个具有诊断价值的基因构建的列线图具备良好的预测性能。