[关键词]
[摘要]
目的 为解决中药有效成分信息缺失、药效物质基础不清楚导致其现代作用机制不明的问题,借助中药药性信息和成分的化学结构信息,利用归纳矩阵填充方法预测中药潜在活性成分。方法 首先,基于中药药性和化学成分信息构建中药-成分关联矩阵;其次,利用中药-成分关联矩阵中潜在的结构信息、中药药性信息和成分的化学结构信息,构建中药相似度矩阵和成分相似度矩阵;最后,基于中药相似度矩阵和成分相似度矩阵填充中药-成分关联矩阵。结果 归纳矩阵填充在中药数据集中使用留一法交叉验证得到的曲线下面积(area under curve,AUC)值为0.768 8。对丹参进行分析,丹参的活性化学成分隐丹参酮、丹参酮IIA、丹参酚醌和丹参醇等分别获得了较高的评分,该预测结果与实际相一致。结论 借助归纳矩阵填充结合中药药性信息和成分的化学结构信息,可有效预测中药的潜在活性成分,为研究中药的现代作用机制提供了新的途径。
[Key word]
[Abstract]
Objective In order to solve the problem of lack of information on the effective ingredients in traditional Chinese medicines (TCMs) and unclear pharmacodynamic material basis, which lead to the unknown modern mechanism of action of TCMs, the inductive matrix filling method was applied to predict the potential active ingredients of TCMs by utilizing the properties information of TCMs and chemical structure information of the ingredients in TCMs. Methods Firstly, the TCM-component association matrix was constructed based on the properties of TCMs and chemical components information. Secondly, the TCM and ingredient similarity matrix were constructed by the potential structural information, TCMs' properties information, and ingredients' chemical structure information in the TCM-component association matrix. Finally, the TCM-component association matrix was filled with the TCM similarity matrix and ingredient similarity matrix. Results The inductive matrix was filled in the herbal dataset, and the area under curve (AUC) value was 0.768 8 through least-one-out cross-validation. Analysis of Danshen (Salviae Miltiorrhizae Radix et Rhizoma) showed that the active chemical components of Salviae Miltiorrhizae Radix et Rhizoma, such as cryptotanshinone, tanshinone IIA, miltionone, and danshenol, received high scores, and the predicted results were consistent with reality. Conclusion By using induction matrix filling and combining TCM's properties information and chemical structure information of the ingredients in TCM, potential active components of TCMs can be effectively predicted, providing a new approach for studying the modern mechanism of action of TCMs.
[中图分类号]
R284;G30
[基金项目]
国家自然科学基金面上项目(82274215);湖南省自然科学基金项目(2023JJ60124);湖南省教育厅重点项目(22A0255,22A0281);湖南省中医药科研重点课题(2023-24);长沙市自然科学基金项目(kq2202265)