In silico prediction of potential drug‐induced nephrotoxicity with machine learning methods

Abstract

In recent years, drug-induced nephrotoxicity has been one of the main reasons for the failure of drug development. Early prediction of the nephrotoxicity for drug candidates is critical to the success of clinical trials. Therefore, it is very important to construct an effective model that can predict the potential nephrotoxicity of compounds. Machine learning methods have been widely used to predict the physico-

Prediction of IDO1 Inhibitors by a Fingerprint‐Based Stacking Ensemble Model Named IDO1Stack

“…All molecules were converted to Simplified Molecular Input Line Entry System (SMILES). To ensure quality and uniqueness, the data were preprocessed by (1) normalizing all molecular structures, (2) removing inorganics, (3) converting salts to the corresponding acids or bases, and (4) filtering out duplicate molecular structures [34] . We randomly divided the processed data into a training set and a test set in an 8 : 2 ratio.…”

“…Where, $${{N}_{total}}$$ is the total number of inhibitor and noninhibitor compounds; N is the number of compounds with IDO1 inhibitory activity; $${{N}_{fragment}}$$ indicates the number of compounds containing a fragment in IDO1 inhibitors; $${{N}_{fragment\_total}}$$ represents the number of compounds containing a fragment in both classes. In addition, we used information gain (IG) and accuracy (ACC) [34] to evaluate and screen dominant substructures. IG was used to measure the importance of substructures to the classification system, and the higher its value is, the higher the importance is.…”

Prediction of the effects of small molecules on the gut microbiome using machine learning method integrating with optimal molecular features

“…The definition of AD is an important consideration for structure‐activity relationship (SAR) modeling according to the OECD guidelines [ 23 ]. The AD of the prediction models means that the model prediction is reliable in this chemical space region.…”

Toxicological data bank bridges the gap between environmental risk assessment and green organic chemical design in One Health world

“…186 Similarly, Gong et al found several substructures with high potential to cause nephrotoxicity by using the SARpy and information gain methods, including four-membered β-lactam combined with a six-membered hydrothiazide ring, 4-propylcyclohexane-1,2-diol, 4-amino- N -propylbutanamide, and phosphoryl group. 187 Shi et al mined some structural alerts for chemical nephrotoxicity by using a consensus model developed by them. According to their findings, phenyl fluoride, polyamine, benzimidazole, and toluene are the high-frequency substructures only present in nephrotoxic chemicals.…”