目的 基于生物信息学、机器学习与分子模拟研究槟榔碱对心肌肥厚的作用机制。方法 基于GEO数据库中与心肌肥厚相关的转录组数据集GSE249925进行差异表达分析。采用3种机器学习算法[最小绝对收缩与选择算子（LASSO）回归、随机森林（RF）、支持向量机-递归特征消除（SVM-RFE）]筛选关键候选基因。通过基因本体论（GO）和京都基因和基因组百科全书（KEGG）富集分析探讨其潜在的生物学功能和信号通路。采用分子对接和分子动力学模拟评估槟榔碱与靶蛋白的结合特性和结构稳定性。结果 共识别出28个可能参与槟榔碱诱导心肌肥厚的候选基因，经过3种机器学习算法联合筛选，最终确定4个核心靶点（STAT3、SFRP4、S100A12、MT1A）。这些基因在心肌肥厚组织中表达显著异常，并具有良好的诊断效能。富集分析显示其涉及多个关键通路，包括Janus激酶信号转导/转录激活因子（JAK/STAT）信号通路、缺氧诱导因子1（HIF-1）信号通路、晚期糖基化终产物-其受体（AGE-RAGE）信号通路、以及磷脂酰肌醇3-激酶/蛋白激酶B（PI3K/Akt）。分子对接及动力学模拟结果表明，槟榔碱可与STAT3和S100A12稳定结合。结论 阐明了槟榔碱诱导心肌肥厚的潜在分子机制，识别出多个关键调控靶点及信号通路，为进一步揭示槟榔毒性作用机制以及开发靶向干预策略提供了理论依据与计算支撑。

Objective To study the mechanism of arecoline effect on myocardial hypertrophy based on bioinformatics, machine learning, and molecular simulation. Methods To conduct differential expression analysis based on the transcriptome dataset GSE249925 related to myocardial hypertrophy in the GEO database. Three machine learning algorithms (LASSO, RF, SVM-RFE) were used to screen key candidate genes. GO and KEGG enrichment analysis, their potential biological functions and signaling pathways were explored. Molecular docking and molecular dynamics simulations were employed to evaluate the binding characteristics and structural stability of arecoline with the target protein. Results A total of 28 candidate genes that might be involved in the induction of cardiac hypertrophy by palmatine were identified. After being jointly screened by three machine learning algorithms, four core targets (STAT3, SFRP4, S100A12, and MT1A) were finally determined. These genes showed significant abnormal expression in the cardiac hypertrophy tissues and had good diagnostic efficacy. Enrichment analysis revealed that they were involved in multiple key pathways, including JAK/STAT pathway, HIF-1 pathway, AGE-RAGE pathway, and PI3K/Akt pathway. Molecular docking and kinetic simulation results indicated that arecoline could stably bind to STAT3 and S100A12. Conclusion This study systematically elucidates the potential molecular mechanisms by which arecoline induces cardiac hypertrophy. Several key regulatory targets and signaling pathways were identified, providing novel insights into the toxicological impact of arecoline and its possible biomarker roles. These findings offer a theoretical and computational basis for future mechanistic research and targeted intervention strategies.

R996.2

广东省中医药局科研项目（ 20231297）；深圳市科技计划项目（ JCYJ20230807141402005）；深圳市“医疗卫生三名工程”资助项目（ SZZYSM202106006） ；深圳市宝安区医疗卫生科研项目 （ 2022JD260） ；深圳市宝安区公立医院高质量发展研究项目（ BAGZL2024068）