Modeling the Biodegradability of Chemical Compounds Using the Online CHEmical Modeling Environment (OCHEM)

Abstract

Biodegradability describes the capacity of substances to be mineralized by free‐living bacteria. It is a crucial property in estimating a compound’s long‐term impact on the environment. The ability to reliably predict biodegradability would reduce the need for laborious experimental testing. However, this endpoint is difficult to model due to unavailability or inconsistency of experimental data. Our approach makes use of the Online Chemical Modeling Environment (OCHEM) and its rich supply of machine learning methods and descriptor sets to build classification models for ready biodegradability. These models were analyzed to determine the relationship between characteristic structural properties and biodegradation activity. The distinguishing feature of the developed models is their ability to estimate the accuracy of prediction for each individual compound. The models developed using seven individual descriptor sets were combined in a consensus model, which provided the highest accuracy. The identified overrepresented structural fragments can be used by chemists to improve the biodegradability of new chemical compounds. The consensus model, the datasets used, and the calculated structural fragments are publicly available at http://ochem.eu/article/31660.

ToxCast EPA in Vitro to in Vivo Challenge: Insight into the Rank-I Model

“…4,22,23,42 Final statistical results for the top-ranking models are summarized in Table 2. Below, we provide a detailed analysis of the steps, which were used to develop the model.…”

Identifying potential endocrine disruptors among industrial chemicals and their metabolites – development and evaluation of in silico tools

“…Obtained statistics for the EAT models with highest balanced accuracy corresponded or were slightly higher than for recently published models using similar data (Li and Gramatica, 2010b;Jensen et al, 2011;Papa et al, 2013;Zhang et al, 2013). The ASNN models are, however, difficult to interpret and thus as suggested by Vorberg and Tetko (2014), PLS models were used to increase our understanding on fundamental structural characteristics of EAT binders and non-binders. These models are described in detail in Section S5 and Table S4.…”

QSAR models and scaffold-based analysis of non-nucleoside HIV RT inhibitors

“…Models with the ASNN method had the best quality and hence were used for further analysis. Studies have shown that the consensus model calculated out of individual models performed best [45,56,57]. An average consensus model was built using 10 QSAR models with the exception of that with spectrophore descriptors, due to its low performance.…”

“…A threshold value of 95% of compounds from the training set was used to determine the qualitative ADs of models. It is assumed that 5% of compounds outside the AD have little effect on the prediction confidence [56]. A Williams"s plot of the consensus model with CONSENSUS-STD as a distance to model is shown in Figure 2, and the AD defined above was used to warn users about unreliable predictions.…”