目的 通过生物信息学技术遴选干预卵巢癌的关键基因并分析其临床价值，结合中医理论预测潜在治疗OV的守正创新中药组方并分析其作用机制。方法 首先，从TCGA数据库下载卵巢癌活检基因表达数据，使用Estimate计算肿瘤微环境中基质得分（stromal score），以中位数为标准分为高、低评分2组，采用Limma包以|log₂(fold change)|＞2、P＜0.05为标准筛选显著差异表达基因（differentially expressed genes，DEGs），通过ClusterProfiler包对DEGs进行基因本体（gene ontology，GO）功能和京都基因与基因百科全书（Kyoto encyclopedia of genes and genomes，KEGG）通路富集分析。DEGs导入STRING数据库中进行蛋白互作网络分析，结果通过Cytoscape中插件MCODE、cytoHubba分析获取关键基因。其次，从生存分析、差异表达、免疫细胞浸润、基因集变异分析（gene set variation analysis，GSVA）评分等角度验证关键基因的临床价值。最后，将关键基因导入Coremine Medical数据库预测具有潜在治疗作用的创新中药组合，基于卵巢癌病机筛选核心中药用药模式及药味作为守正中药组合，整合2部分作为守正创新中药方，采用中药复方网络药理学的方法分析其作用机制。结果 获取卵巢癌患者活检数据381例，以中位基质得分（−312.742 068 5）分为低（190例）、高（191例）2组，筛选得到DEGs 202个。GO富集分析显示DEGs主要干预体液免疫应答、补体激活、吞噬功能等生物过程；免疫球蛋白复合物、质膜组成、T细胞受体复合物等细胞组成；抗原结合、免疫球蛋白受体结合、整合素结合等分子功能。KEGG富集分析显示DEGs主要干预信号分子和相互作用、免疫系统、运输和分解代谢等功能信号通路。蛋白互作及MCODE、cytoHubba筛选得V型胶原α1（collagen type V alpha 1，COL5A1）、纤维连接蛋白（fibronectin，FN1）、XI型胶原α1（collagen type XI alpha 1，COL11A1）、I型胶原α2（collagen type I alpha 2，COL1A2）、原纤蛋白1（fibrillin 1，FBN1）、I型胶原α1（collagen type I alpha 1，COL1A1）、核心蛋白聚糖（decorin，DCN）、V型胶原α2（collagen type Ⅴ alpha 2，COL5A2）、III型胶原α1（collagen type Ⅲ alpha 1，COL3A1）、基膜聚糖（lumican，LUM）10个关键基因。临床价值分析显示，关键基因表达显著影响卵巢癌患者无进展生存期、总生存期、进展后生存期。与正常群体相比，LUM、COL5A1、COL5A2、DCN、FBN1表达显著增高，COL11A1显著降低，主成分分析显示关键基因表达可将正常人和卵巢癌患者分为2组。关键基因-GSVA-免疫细胞浸润关联分析显示，关键基因表达、GSVA评分与免疫细胞浸润评分显著相关且与多种免疫细胞均显著相关。预测得创新部分中药组合为南木香、千层塔、茺蔚子、猫爪草、杜仲、肿节风、红娘子、预知子、堇菜、鹿角，守正部分中药组合为川芎、甘草、半夏、黄芩、当归、熟地黄、桔梗、大黄、茯苓。网络药理学分析显示，守正创新中药方包含化合物2 199个，对应去重后靶蛋白1 749个，与DEGs交集34个。GO和KEGG富集分析显示交集靶蛋白主要通过干预白细胞分化、血管生成、白细胞趋化等生物过程；胶原蛋白XI型三聚体、含胶原蛋白的细胞外基质等细胞组成；血小板衍生生长因子结合、趋化因子活性、糖胺聚糖结合等分子功能；糖尿病并发症中的晚期糖基化终产物-晚期糖基化终产物受体（advanced glycation end products-receptor for advanced glycosylation end products，AGE-RAGE）、白细胞跨内皮迁移、血管平滑肌收缩、癌症蛋白多糖、蛋白质的消化和吸收、松弛素、ECM与受体的相互作用等信号通路表达作用于肿瘤微环境。结论 基质细胞与卵巢癌发生发展过程中免疫应答、物质转运、耐药、能量代谢等生物过程密切相关，靶向基质细胞为卵巢癌治疗提供可行策略，中药“多成分、多靶点、多机制”的作用特性与其相适应，为靶向肿瘤微环境的现代中药复方开发提供思路。

Objective Through bioinformatics methods, we aim to screen out the key genes involved in the intervention of ovarian cancer (OV) and analyze their clinical value. Furthermore, based on traditional Chinese medicine (TCM) theory, we predict potential therapeutic herbal formulas and investigate their underlying mechanisms. Methods Download OA biopsy gene expression data from the TCGA database, use Estimate to calculate the stromal score in the tumor microenvironment, and divide it into high and low score groups using the median as the standard. Use the Limma package to calculate|log₂(fold change)|> 2, P< 0.05 was used as the standard to screen significantly differentially expressed genes (DEGs), and the DEGs were enriched using the ClusterProfiler package for gene ontology (GO) enrichment analysis and Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis. DEGs were imported into the STRING database for protein interaction analysis, and the results were imported into cytoscape using the plug-in MCODE and cytoHubba plug-ins to analyze and intersect to obtain key genes. The key genes’ clinical value were verified from the perspectives of survival analysis, differential expression, immune cell infiltration, and gene set variation analysis (GSVA) scores. Key genes were introduced into the Coremine Medical database to predict innovative TCM combinations with potential therapeutic effects. Secondly, based on the pathogenesis of ovarian cancer, core TCM medication patterns and medicinal flavors were screened as a TCM combination. The two parts of the formula were integrated and the new formula was analyzed using the method of traditional Chinese medicine compound network pharmacology. mechanism of action. Results A total of 381 cases of OV patient biopsy data were downloaded, which were divided into two groups: low (190 cases) and high (191 cases) based on the median stromalscore (−312.742 068 5), 202 DEGs were screened. GO enrichment analysis showed that DEGs mainly interfere with biological processes such as humoral immune response, complement activation, and phagocytosis, as well as cellular components such as immunoglobulin complexes, plasma membrane components, and T cell receptor complexes, as well as antigen binding and immunoglobulin receptor binding, integrin binding and other molecular functions. KEGG enrichment analysis showed that DEGs mainly interfere with functional signaling pathways such as signaling molecules and interactions, immune system, transport, and catabolism. Protein interaction, MCODE, and cytoHubba screening resulted in collagen type V alpha 1 (COL5A1), fibronectin (FN1), collagen type XI alpha 1 (COL11A1), and type I collagen α2 (COL1A2), fibrillin 1 (FBN1), type I collagen α1 (COL1A1), decorin (DCN), collagen type V alpha 2 (COL5A2), collagen type III alpha 1 (COL3A1), and Lumican (LUM). Clinical value analysis shows that the expression of key genes significantly affects the progression-free survival, overall survival, and post-progression survival of patients with ovarian cancer. Compared with the normal group, the expression ofLUM,COL5A1, COL5A2,DCN, andFBN1 was significantly increased, and COL11A1 was significantly decreased. Principal component analysis showed that the expression of key genes can divide normal and ovarian cancer into two groups. Key gene-GSVA-immune cell infiltration correlation analysis showed that key gene expression, GSVA score and were significantly correlated with infiltration scoreand and significantly correlated with a variety of immune cells. The predicted innovative combinations of TCM are Nanmuxian晧?伨嘼?￣呁慲物杳整瑯楬湯杣?瑩桡攠?獵瑮牮潡浮慥汮?捩敳氼氯獩￣牆敲灡牮散獨攮温琼獩￣愬?瘯楩愾戼汩放?琼栯敩爾慑灩敡畮瑣楥据?獴瑡爠愨琼敩朾祌?晣潯牰?佤噩?瑭爠敳慥瑲浲敡湴瑵??启桩放?畨湵楮煢甮攩?挠桃慨牯慮捧瑷敥物楺獩琠椨挼獩￣潍景?周????楲湴挠汆畲摵楩湴朼?楩琾猩?洠畍污瑯楺?捵潡浣灡潯渠攨渼瑩??浡畮汵瑮楣?瑬慵牳朠整瑥??慡湴摵?洼甯汩琾椩?洠敄捵桺慨湯楮獧洠?愼灩瀾牅潵慣捯桭??慡牥攠?睯敲汴汥?猼甯楩琾攩搬?晚潨牯?瑧桪楩獥?灥畮牧瀠漨猼敩??呡桲散?灮牤敲獡敥渠瑈?獲瑢畡搼礯?瀾爩漬瘠楈摯敮獧?楩湡獮楧杺桩琠猨?楩渾瑈潵?瑣桨敹?瀠潳瑡敮湧瑵楩慮汥?搼支癩放汄潥瀠浇敥湥瑲?漼晩￣浈潵摥散牨湹?吠???晲潡牣浩畣污愼琯楩漾湄獩?瑴桡慮瑴?琼慩爾杈敵瑥?瑨桹敳?瑰畨浩潬牡?浭楡捴牡漼支湩瘾楆牡潢湲浩散湩瑵??漬映晙敵牺楨湩杺?渠敛眼?搾楁牫敥换瑩楡漠湱獵?普潡牴?琼栯敩￣攨硈灯汵潴牴愮琩椠潄湥?潮晥?琩桝攬爠慊灩敮畣瑡楩挠?漼灩琾楖潩湯獬?椠湶?潲癥慣牵楮慤湡?振慩渾捁攮爠?ray),Lujiao (Cervi Cornu) and the traditional Chinese medicine combinations are Chuanxiong (Chuanxiong Rhizoma), Gancao (Glycyrrhizae Radix et Rhizoma), Banxia (Pinelliae Rhizoma), Huangqin (Scutellariae Radix), Danggui (Angelicae Sinensis Radix), Shudihuang (Rehmanniae Radix Praeparata), Jiegeng (Platycodonis Radix), Dahuang (Rhei Radix et Rhizoma), Fuling (Poria). Network pharmacology analysis showed that the new formula contains 2 199 compounds, corresponding to 1 749 target proteins after deduplication, and 34 intersections with DEGs. GO and KEGG enrichment analysis showed that the intersection target proteins mainly interfere with biological processes such as leukocyte differentiation, angiogenesis, and leukocyte chemotaxis; collagen type XI trimers, collagen-containing extracellular matrix and other cellular components; platelet-derived growth factor binding, chemokine activity, glycosaminoglycan binding and other molecular functions; advanced glycation end products-receptor for advanced glycosylation end products in diabetic complications, leukocyte transendothelial migration, vascular smooth muscle contraction, cancer proteoglycans, protein digestion and absorption, relaxin, ECM and receptors The interaction and expression of signaling pathways act on the tumor microenvironment. Conclusion The stromal cells are intimately involved in biological processes such as immune response, material transport, drug resistance, and energy metabolism during the development and progression o

Q811.4;TP18;R285

上海医药中药传承和创新平台能力建设（2020006）