傳統中藥的處方由於來自天然物,而常被認為是安全的療方,因此在西方社會中也日漸普及地用於治療和預防疾病。然而傳統中藥在使用上往往缺乏科學化的安全性評估。電腦輔助篩選系統在評估藥物安全性上較傳統實驗省時、省成本,且篩選得到的結果也有良好的再現性,欲藉此方式篩出傳統中藥當中潛在具有肝毒性的成分。首先調整肝毒性知識庫的品項建立一套定量構效關係模型。這是整合臨床試驗或上市後藥物監視系統中會誘發肝損傷的藥物及相關資訊的資料集。再經過巢狀10層交叉驗證及外部驗證皆得到良好的結果。最後應用此一模型掃描TCM Database@Taiwan — 全世界最大的傳統中藥資料庫,篩選出總共6853個成分(21.4%)具有潛在的肝毒性。進一步將肝毒性預測值最高的前100名化合物進行文獻佐證。研究顯示傳統中藥成分的毒性證據並不充足,並證實此一定量構效關係模型可以作為輔助的工具來評估傳統中藥的安全性。
The perception that natural substances are deemed safe has made traditional Chinese medicine (TCM) gain popularity in the treatment and prevention of disease globally. However, such an assumption is often misleading for lack of scientific validation. To assess the safety of TCM, the in silico screening provides major advantages over the classical laboratory approaches in resource- and time-saving and full reproducibility. To screen the hepatotoxic potential of the active compounds of TCM, a quantitative structure-activity relationship model was firstly established by utilizing chemicals from the Liver Toxicity Knowledge Base. These chemicals were annotated with drug-induced liver injury information obtained from clinical trials and post-marketing surveillance. The performances of nested 10-fold cross-validation were 79.1%, 91.2%, and 53.8% for accuracy, sensitivity, and specificity, respectively. The external validation on 84 known hepatotoxic ingredients of common herbal medicines yielded a high accuracy (92.8%). After screening the TCM Database@Taiwan, the world’s largest TCM database, a total of 6853 (21.4%) ingredients were predicted with hepatotoxic potential. One hundred chemicals predicted with highest hepatotoxic potential by our model were further verified by published literatures. Our study indicates this model can serve as a complementary tool to evaluate the safety of TCM.