Wu Leihong, Gao Xiumei, Wang Linli, Liu Qian, Fan Xiaohui, Wang Yi, Cheng Yiyu
Department of Chinese Medicine Science & Engineering, Zhejiang University, Hangzhou 310058, China.
Zhongguo Zhong Yao Za Zhi. 2011 Nov;36(21):2907-10.
To predict multi-targets by multi-compounds found in Aconiti Lateralis Radix Praeparata and construct the corresponding multi-compound-multi-target network.
Based on drug-target relationships of FDA approved drugs, a model for predicting targets was established by random forest algorithm. This model was then applied to predict the targets of Aconiti Lateralis Radix Praeparata and construct the multi-compound-multi-target network.
The predicted targets of 22 compounds of Aconiti Lateralis Radix Praeparata are validated by literature. Each compound in the established network was correlated with 16. 3 targets on average, while each target was correlated with 4. 77 compounds on average, which reflects the "multi-compound and multi-target" characteristic of Chinese medicine.
The proposed approach can be used to find potential targets of Chinese medicine.
预测制附子中多种化合物的多靶点,并构建相应的多化合物-多靶点网络。
基于美国食品药品监督管理局(FDA)批准药物的药物-靶点关系,采用随机森林算法建立靶点预测模型。然后将该模型应用于预测制附子的靶点并构建多化合物-多靶点网络。
制附子22种化合物的预测靶点经文献验证。所建立网络中的每种化合物平均与16.3个靶点相关,而每个靶点平均与4.77种化合物相关,这体现了中药“多成分、多靶点”的特点。
所提出的方法可用于发现中药的潜在靶点。