A novel prediction approach for antimalarial activities of Trimethoprim, Pyrimethamine, and Cycloguanil analogues using extremely randomized trees

详细信息    查看全文

• 作者：Cholwich Nattee^a ; ; Nirattaya Khamsemanan^a ; ^{nirattaya@siit.tu.ac.th} ; Luckhana Lawtrakul^a ; ; Pisanu Toochinda^a ; Supa Hannongbua^b
• 关键词：Extremely randomized trees ; Decision trees ; Structure&ndash ; activity relationship ; Molecular modeling ; Antimalarials ; Dihydrofolate reductase ; Trimethoprim ; Pyrimethamine ; Cycloguanil ; Hydrophobic feature
• 刊名：Journal of Molecular Graphics and Modelling
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：71
• 期：Complete
• 页码：13-27
• 全文大小：1354 K
• 卷排序：71

文摘

First combined QSAR model for antimalarial activities of TMP, PYR, CYC is proposed. The prediction model, using extremely randomized trees, yields a strong R2 of 0.996. The proposed model outperforms other existing techniques. The proposed model reveals common conditions associated with low predicted Ki. Compounds with ASA H > 575.80 and Log S = 4.36 have predicted Ki lower than 10.