Developing an artificial intelligence method for screening hepatotoxic compounds in traditional Chinese medicine and Western medicine combination

• DOI: 10.1186/s13020-022-00617-4
• Publication Date: 2022-05-17T04:02:52Z
• Authors: Zhao Chen; Mengzhu Zhao; Liangzhen You; Rui Zheng; Yin Jiang; Xiaoyu Zhang; Ruijin Qiu; Yang Sun; Haie Pan; Tianmai He; Xuxu Wei; Zhineng Chen; Chen Zhao; Hongcai Shang
• Keywords: Data set; Deep belief network; AdaBoost

Traditional Chinese medicine and Western combination (TCM-WMC) increased the complexity of compounds ingested.To develop a method for screening hepatotoxic in TCM-WMC based on chemical structures using artificial intelligence (AI) methods.Drug-induced liver injury (DILI) data was collected from public databases published literatures. The total dataset formed by DILI randomly divided into training set test at ratio 3:1 approximately. Machine learning models SGD (Stochastic Gradient Descent), kNN (k-Nearest Neighbor), SVM (Support Vector Machine), NB (Naive Bayes), DT (Decision Tree), RF (Random Forest), ANN (Artificial Neural Network), AdaBoost, LR (Logistic Regression) one deep model (deep belief network, DBN) were adopted to construct compounds.Dataset 2035 this research, which 1505 as 530 set. Results showed that obtained 0.838 classification accuracy (CA), 0.827 F1-score, 0.832 Precision, Recall, 0.814 area under curve (AUC) 0.767 CA, 0.731 F1, 0.739 AUC set, better than other eight machine methods. DBN 82.2% higher any set.The AI expected effectively screen TCM-WMC.