目的 采用数据挖掘、网络药理学和分子对接技术，归纳分析中药复方治疗结肠癌的用药规律、性质及作用机制，并通过实验验证相关结果。方法 检索中国学术期刊全文数据库（CNKI）、万方数据库（Wanfang Data）、维普生物医学数据库（VIP）、PubMed、Web of Science数据库获取，中药复方治疗结肠癌的临床研究相关文献，根据纳入和排除标准进行筛选，去重后构建有效治疗结肠癌的中药复方集，对中药频次、功效、性味、归经进行挖掘，并运用SPSS Modeler 18.0和SPSSStatistics 27.0进行关联规则分析和聚类分析，得到核心中药组合；通过中药系统药理学数据库与分析平台（TCMSP）检索活性成分和作用靶点，通过GeneCards、OMIM和TTD数据库检索疾病靶点，利用Venny 2.1.0平台获取药物与结肠癌的交集靶点，通过STRING数据库构建蛋白质-蛋白质相互作用（PPI）网络，运用Cytoscape软件建立“药物-成分-靶点”可视化网络图，通过DAVID在线分析平台进行基因本体（GO）注释及京都基因与基因组百科全书（KEGG）通路富集分析，并使用CB-Dock2平台进行分子对接及可视化处理。通过CCK-8实验检测活性成分对细胞增殖的影响；流式细胞仪检测活性成分对细胞周期的影响；实时荧光定量PCR法（qRT-PCR）检测活性成分对关键蛋白靶点分子表达的影响。结果 共纳入87个处方，包含194味中药，其中白术、茯苓、黄芪、党参和薏苡仁的使用频次较高。中药功效以补虚药、利水渗湿药和清热药为主。药性以温、平为主，药味以甘、苦、辛居多，主要归脾、肺和胃经。关联规则分析共得到64个药物组合，聚类分析得到4类中药。核心中药白术、茯苓、黄芪、党参和薏苡仁与结肠癌的交集靶点131个，核心靶点有肿瘤蛋白53（TP53）、丝氨酸/苏氨酸激酶1（AKT1）、Jun原癌基因（JUN）等，核心成分有槲皮素、木犀草素、山柰酚等，主要涉及癌症通路、白细胞介素-17、磷脂酰肌醇3激酶（PI3K）-Akt、晚期糖基化终末产物-晚期糖基化终产物受体（AGE-RAGE）信号通路等。分子对接发现核心成分与核心靶点对接活性良好，JUN与槲皮素的结合强度最高。实验研究表明，槲皮素可抑制结肠癌细胞HT-29细胞增殖、阻滞其细胞周期，抑制JUN基因表达水平。结论 该研究归纳总结了中药复方治疗结肠癌的用药规律，多维数据挖掘分析证实治疗结肠癌的中药处方大多具有健脾益气、滋补肝肾和活血化瘀的特性，其关键活性成分槲皮素、木犀草素、山柰酚主要作用于TP53、AKT1、JUN等结肠癌靶点，进一步实验表明槲皮素是治疗结肠癌的潜在候选化合物。

Objective To summarize the medication rules and properties of traditional Chinese medicine (TCM) formulas in treating colon cancer and explore their mechanisms of action through data mining, network pharmacology, and molecular docking techniques. The results were verified through experiments. Methods Relevant clinical research literature on TCM formulas for treating colon cancer was retrieved from databases such as CNKI, Wanfang Data Knowledge Service Platform, VIP Chinese Journal Service Platform, PubMed, and Web of Science. After screening based on inclusion and exclusion criteria and removing duplicates, a set of effective TCM formulas for treating colon cancer was constructed. The frequency, efficacy, taste, and meridian tropism of the herbs were mined, and association rule analysis and cluster analysis were conducted using SPSS Modeler 18.0 and SPSS Statistics 27.0 to obtain core herb combinations. Active components and target proteins were retrieved from the TCMSP database, and disease targets were retrieved from GeneCards, OMIM, and TTD databases. The intersection targets of drugs and colon cancer were obtained using the Venny 2.1.0 platform, and a protein-protein interaction (PPI) network was constructed using the STRING database. A “drug-component-target” visualization network was established using Cytoscape software, and Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed using the DAVID online analysis platform. Molecular docking and visualization were conducted using the CB-Dock2 platform. Finally, the effects of active components on cell proliferation were detected by CCK-8 assay; the effects on cell cycle were detected by flow cytometry; And the effects on the expression of key protein target molecules were detected by real-time fluorescence quantitative PCR (qRT-PCR). Results A total of 87 prescriptions were included, containing 194 TCM herbs. Among them, Atractylodes macrocephala, Poria cocos, Astragalus membranaceus, Codonopsis pilosula, and Coicis Semen were frequently used. The main efficacy of the herbs was tonifying, promoting diuresis and eliminating dampness, and clearing heat. The nature of the herbs was mainly warm and neutral, and the taste was mainly sweet, bitter, and pungent, mainly affecting the spleen, lung, and stomach meridians. Association rule analysis yielded 64 drug combinations, and cluster analysis identified four categories of TCM herbs. The intersection targets of the core herbs Atractylodes macrocephala, Poria cocos, Astragalus membranaceus, Codonopsis pilosula, and Coicis Semen and colon cancer were 131, with core targets including tumor protein 53 (TP53), serine/threonine kinase 1 (AKT1), Jun proto-oncogene (JUN), etc. The core components were quercetin, luteolin, kaempferol, etc., mainly involving cancer pathways, interleukin-17, PI3K-Akt, AGE-RAGE signaling pathways, etc. Molecular docking revealed good binding activity between core components and core targets, with the highest binding strength between JUN and quercetin. Experimental studies showed that quercetin could inhibit the proliferation of colon cancer HT-29 cells, arrest their cell cycle, and suppress the expression level of the JUN gene. Conclusion This study summarized the medication rules of TCM formulas in treating colon cancer. Multi-dimensional data mining analysis confirmed that most TCM prescriptions for treating colon cancer have the characteristics of tonifying the spleen and qi, nourishing the liver and kidney, and promoting blood circulation and removing blood stasis. The key active components quercetin, luteolin, and kaempferol mainly act on colon cancer targets such as TP53, AKT1, and JUN. Further experiments indicated that quercetin is a potential candidate compound for treating colon cancer.

R975

国家自然科学基金资助项目（82274680）；江西省重点研发计划课题（20243BCC31011）；江西省自然科学基金（20262BAC200343）；经典名方现代中药创制全国重点实验室自主部署项目（2026QZZZBSXM005）；江西省中医药管理局科技计划项目（2023B1224）；江西中医药大学博士科研启动基金（2022BSZY013，2025WBZR002）；江西中医药大学大学生创新创业训练计划项目（202610412315）