In Silico Assessment of Chemical Biodegradability

Abstract

Biodegradation is the principal environmental dissipation process. Due to a lack of comprehensive experimental data, high study cost and time-consuming, in silico approaches for assessing the biodegradable profiles of chemicals are encouraged and is an active current research topic. Here we developed in silico methods to estimate chemical biodegradability in the environment. At first 1440 diverse compounds tested under the Japanese Ministry of International Trade and Industry (MITI) protocol were used. Four different methods, namely support vector machine, k-nearest neighbor, naïve Bayes, and C4.5 decision tree, were used to build the combinatorial classification probability models of ready versus not ready biodegradability using physicochemical descriptors and fingerprints separately. The overall predictive accuracies of the best models were more than 80% for the external test set of 164 diverse compounds. Some privileged substructures were further identified for ready or not ready biodegradable chemicals by combining information gain and substructure fragment analysis. Moreover, 27 new predicted chemicals were selected for experimental assay through the Japanese MITI test protocols, which validated that all 27 compounds were predicted correctly. The predictive accuracies of our models outperform the commonly used software of the EPI Suite. Our study provided critical tools for early assessment of biodegradability of new organic chemicals in environmental hazard assessment.

In silico Prediction of Drug Induced Liver Toxicity Using Substructure Pattern Recognition Method

“…Salts were converted into the corresponding acids or bases; water molecules were removed from hydrates. The last step was to remove duplicates and kept the molecules with molecular weight higher than 40 and lower than 800 in the data set 21–22. Finally, a large diverse DILI database containing 1229 unique molecules was obtained.…”

In silico prediction of chemical reproductive toxicity using machine learning

“…If a substructure was more frequently presented in a positive reproductive toxicity chemical class, this substructure was called an SA involved in potential reproductive toxin. The “frequency of a fragment” enrichment factor in the toxic chemical was defined as follows (Cheng et al, ). where is the number of compounds containing the fragment in the positive reproductive toxicity chemicals, N toatal is the total number of compounds, N fregment _ total is the total number of compounds containing this fragment and N B is the number of compounds in the positive reproductive toxicity chemical class.…”

“…Therefore, it is necessary to define the AD of models. In this study, we conducted a similarity‐based AD analysis (Cheng et al, ) by comparing the distance between a query chemical and its kNNs in the training set (Equation ) with a predefined AD threshold, D T . where is the average Euclidean distance of each compound in the training set and its nearest neighbor, σ is the standard deviation of these Euclidean distances and Z an arbitrary value varying from −1 to −4 in this study.…”

Computational QSAR model combined molecular descriptors and fingerprints to predict HDAC1 inhibitors

“…Here, we used fingerprints as attributes for QSAR modeling based on high-quality diverse data from the literature and databases. Fingerprints were good methods for chemical toxicity prediction and chemical metabolic property prediction which always made a direct connection between the chemical structure and toxicity endpoint of global compounds [33,34]. In our study, we distinguish novel HDAC1 inhibitors and non-inhibitors by five machine learning methods combined with seven fingerprints.…”