ABSTRACT
Mutagenicity is considered an important endpoint from the regulatory, environmental and medical points of view. Due to the wide number of compounds that may be of concern and the enormous expenses (in terms of time, money, and animals) associated with rodent mutagenicity bioassays, this endpoint is a major target for the development of alternative approaches for screening and prediction. The majority of old-aged expert systems and quantitative structure-activity relationship (QSAR) models may show reduced performance over time for their application on newer chemical candidates; thus, researchers constantly try to improve the modeling strategies. In our report, we initially performed traditional classification-based linear discriminant analysis (LDA) QSAR modeling using the benchmark Ames dataset of diverse chemicals (6512 compounds) to recognize the relationship between the molecules and their potential mutagenic behavior. The classical LDA QSAR model is developed from a selected set of 2D descriptors. The LDA QSAR model was developed by using a total of 31 descriptors identified from the analysis of the most discriminating features. Additionally, we have used similarity-derived features obtained from the read-across (RA) to develop an RA-based QSAR model. The developed RA-based LDA QSAR model has better predictivity, transferability, and interpretability compared to the LDA QSAR model, and it uses a very small number of descriptors compared to the classical QSAR model. Different machine learning (ML) models were also developed using the descriptors appearing in the read-across-based LDA QSAR model for comparative studies. We have checked the prediction quality of 216 true external set compounds using the novel similarity-derived RA model. The performance of the OECD toolbox is also compared with the RA-derived LDA QSAR model for a true external set. The current study aimed to explore the significance of the read-across-based algorithm and its application to the most current experimental mutagenicity data to complement already available expert systems.
Subject(s)
Mutagens , Quantitative Structure-Activity Relationship , Mutagens/toxicity , Expert Systems , Algorithms , Machine LearningABSTRACT
Octanol-water partition coefficient (logKow) and soil organic carbon content normalized sorption coefficient (logKoc) values are two important physicochemical properties in the context of bioaccumulation and environmental fate of organic compounds and their environmental risk assessment. Simple, interpretable and easy-to-derive extended topochemical atom (ETA) indices obtained from 2D structural representation of compounds were used for quantitative structure-property relationship (QSPR) modeling of these two endpoints. Linear regression based models developed using only ETA indices show encouraging statistical and validation results. Based on the information obtained from developed QSPR models, we may conclude that molecular volume, branching pattern, presence of hydrophobic Cl atoms, cyclicity/fusion, polar environment, electron density, unsaturation content, hydrogen bonding propensity or hydrogen bond donor atoms, local topology, presence of heteroatoms and aromaticity are crucial factors in controlling the logKow and logKoc values of the compounds. The suggested explanatory features for different classes of chemicals or the whole diverse set can help in safer designing of chemicals, which is one of the primary agenda of the "Green Chemistry" program.
Subject(s)
Organic Chemicals/toxicity , Quantitative Structure-Activity Relationship , Soil Pollutants/toxicity , Carbon , Linear Models , Octanols/chemistry , Organic Chemicals/chemistry , Soil/chemistry , Soil Pollutants/chemistry , Water/chemistryABSTRACT
Environmental transformation products of pesticides (ETPPs) have a great deal of ecological impact owing to their ability to cause toxicity to the aquatic organisms, which can then be translated to the humans. The limited experimental data on biochemical and toxic effects of ETPPs, the high test costs together with regulatory limitations and the international push to reduce animal testing encourage greater dependence on predictive in silico techniques like quantitative structure-activity relationship (QSAR) models. The aim of the present work was to explore the key structural features, which regulate the toxicity towards fishes, for 85 ETPPs using a partial least squares (PLS) regression based chemometric model developed according to Organisation for Economic Co-operation and Development (OECD) guidelines. The model was extensively validated using both internal and external validation metrics, and the results so obtained justify the reliability and usefulness of the developed model (Q2 = 0.648, R2pred or Q2F1 = 0.734 and Q2F2 = 0.733). From the developed model, we can conclude that lipophilicity, polarity, presence of branching and the functional form of O-atom in the transformed structures of pesticides are the important features that are to be considered during ecotoxicity assessment of ETPPs. The information obtained from the descriptors of the developed model could be utilized in the future for assessing ETPPs with the benefit of providing an early warning of their potentially detrimental effect on fishes for regulatory purposes.