Targeting against HIV/HCV Co-infection using Machine Learning-based multitarget-quantitative structure-activity relationships (mt-QSAR) Methods

Abstract

24Co-infection between HIV-1 and HCV is common today in certain populations. However, treatment of 25 co-infection is full of challenges with special consideration for potential hepatic safety and drug-drug 26 interactions. Multitarget inhibitors with less toxicity may provide a promising therapeutic strategy for 27 HIV/HCV co-infection. However, identification of one molecule acting on multiple targets simultaneously by 28 experimental evaluation is costly and time-consuming. In silico target prediction tools provide more 29 opportunities for the development of multitarget inhibitors. In this study, by combining naive Bayesian (NB) 30 and support vector machine (SVM) algorithms with two types of molecular fingerprints (MACCS and ECFP6), 31 60 classification models were constructed to predict the active compounds toward 11 HIV-1 targets and 4 32 HCV targets based on the multitarget-quantitative structure-activity relationships (mt-QSAR). 5-fold cross-33 validation and test set validation was performed to confirm the performance of 60 classification models. Our 34 results show that 60 mt-QSAR models appeared to have high classification accuracy in terms of ROC-AUC 35 values ranging from 0.83 to 1 with a mean value of 0.97 for HIV-1 models, and ROC-AUC values ranging 36 from 0.84 to 1 with a mean value of 0.96 for HCV. Furthermore, the 60 models were applied to 37 comprehensively predict the potential targets for additional 46 compounds including 27 approved HIV-1 drugs, 38 10 approved HCV drugs and 9 selected compounds known to be active on one or more targets of HIV-1 or 39 those of HCV. Finally, 18 hits including 7 HIV-1 approved drugs, 4 HCV approved drugs and 7 compounds 40 were predicted to be HIV/HCV co-infection multitarget inhibitors. The reported bioactivity data confirmed 41 that 7 compounds actually interacted with HIV-1 and HCV targets simultaneously with diverse binding 42 affinities. Of those remaining predicted hits and chemical-protein interaction pairs involving the potential 43 ability to suppress HIV/HCV co-infection deserve further investigation by experiments. This investigation 44 shows that the mt-QSAR method is available to predict chemical-protein interaction for discovering 45 multitarget inhibitors and provide a unique perspective on HIV/HCV co-infection treatment. 3 46 48 syndrome (AIDS), a pandemic disease[1]. In addition, hepatitis C virus (HCV) infection causes acute and 49 chronic liver disease, including cirrhosis and hepatocellular carcinoma[2]. Unfortunately, since HIV-1 and 50 HCV have similar routes of transmission, the risk of HIV/HCV co-infection is very high[3]. According to the 51 World Health organization (WHO), about 2.3 million persons of the estimated 36.7 million living with HIV 52 had been infected with HCV globally in 2015. It is necessary for those people to be diagnosed and provided 53 with effective and reasonable treatment for both HIV and hepatitis as a priority[4]. However, in the treatment 54 of HIV/HCV co-infection, special considerations must be ...

Activity Prediction of Small Molecule Inhibitors for Antirheumatoid Arthritis Targets Based on Artificial Intelligence

“…At present, with the development of computing power and the accumulation of pharmacological data, artificial intelligence has made great progress in various fields of drug design. − For instance, support vector machine (SVM) as a widely used machine learning method has shown high yield and low false-hit rate in single-target drug screening. , Yabuuchi et al successfully predicted inhibitors of GPCR and kinase targets using SVM and validated their prediction by experiments. Extreme gradient boosting (XGBoost) is an effective and efficient machine learning method in QSAR fields. − XGBoost not only reduces model complexity to prevent overfitting, but also supports parallel processing to reduce computation.…”

“…22−25 For instance, support vector machine (SVM) as a widely used machine learning method has shown high yield and low falsehit rate in single-target drug screening. 26,27 Yabuuchi et al 28 successfully predicted inhibitors of GPCR and kinase targets using SVM and validated their prediction by experiments. Extreme gradient boosting (XGBoost) is an effective and efficient machine learning method in QSAR fields.…”