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

Abstract

The ToxCast EPA challenge was managed by TopCoder in Spring 2014. The goal of the challenge was to develop a model to predict the lowest effect level (LEL) concentration based on in vitro measurements and calculated in silico descriptors. This article summarizes the computational steps used to develop the Rank-I model, which calculated the lowest prediction error for the secret test data set of the challenge. The model was developed using the publicly available Online CHEmical database and Modeling environment (OCHEM), and it is freely available at . Surprisingly, this model does not use any in vitro measurements. The logic of the decision steps used to develop the model and the reason to skip inclusion of in vitro measurements is described. We also show that inclusion of in vitro assays would not improve the accuracy of the model.

Biological activities of extracts from Aspidosperma subincanum Mart. and in silico prediction for inhibition of acetylcholinesterase

“…The overall idea was to build several independent models and average their output. This procedure is known to avoid overfitting and minimize prediction errors (Novotarskyi et al, ). In our approach, 30 independent models were produced through stratified random partition of the original dataset to produce two subsets, the training set and the validation set (Figure ).…”

“…The overall idea was to build several independent models and average their output. This procedure is known to avoid overfitting and minimize prediction errors (Novotarskyi et al, 2016).…”

Transformer-CNN: Fast and Reliable Tool for QSAR

“…Lipo Lipophilicity [47] 4,200 BBBP Blood-brain barrier [43] 2,039 BACE IC50 of human β-secretase 1 (BACE-1) inhibitors [43] 1,513 JAK3 Janus kinase 3 inhibitor [48] 886 DHFR Dihydrofolate reductase inhibition [49] 739 BioDeg Biodegradability [50] 1,737 LEL Lowest effect level [51] 483 RP AR Endocrine disruptors [52] 930 To validate the model, we sampled 500,000 ChEMBL-like SMILES (only 8,617 (1.7%) of them were canonical) from a generator [53] and checked how accurately the model can restore canonical SMILES for these molecules. We intentionally selected the generated SMILES keeping in mind possible application of the proposed method in the artificial intelligence-driven pipelines of de-novo development of new drugs.…”

Public (Q)SAR Services, Integrated Modeling Environments, and Model Repositories on the Web: State of the Art and Perspectives for Future Development

“…Models available on OCHEM were frequently top‐performing within different benchmarking exercises, e.g., prediction of metal complexation, environmental toxicity, readily biodegradability, endocrine disruptors, logP,,, AMES toxicity, CYP450 inhibition, etc., as well as contributed the top‐rank model for the EPA ToxCast and the best overall balanced accuracy for twelve targets of the NIH Tox21 challenges. The OCHEM is widely used for educational purposes by a number of Universities across the world.…”