Predictive Quantitative Structure Toxicity Relationship Study on Avian Toxicity of Some Diverse Agrochemical Pesticides by Monte Carlo Method: QSTR on Pesticides

Predictive Quantitative Structure Toxicity Relationship Study on Avian Toxicity of Some Diverse Agrochemical Pesticides by Monte Carlo Method: QSTR on Pesticides

Amit Kumar Halder (Natural Science Laboratory, Division of Medicinal and Pharmaceutical Chemistry, Department of Pharmaceutical Technology, Jadavpur University, Kolkata, India), Achintya Saha (Department of Chemical Technology, University of Calcutta, Kolkata, India) and Tarun Jha (Natural Science Laboratory, Division of Medicinal and Pharmaceutical Chemistry, Department of Pharmaceutical Technology, Jadavpur University, Kolkata, India)
DOI: 10.4018/IJQSPR.2017010102

Abstract

Application of pesticides may have serious adverse consequences in environment. Birds are one of the most important non-target species that are harmed by agricultural chemical pesticides. In the current study, Monte Carlo optimization based Quantitative Structure Toxicity Relationship (QSTR) analyses were performed on a dataset containing diverse chemical pesticides with toxicity data determined on Bobwhite quail. Hybrid models containing SMILES and graph based descriptors were developed on three different training and test set combinations. The best model was selected based on validation statistics on internal training (n = 96) and external test (n = 31) as well as validation (n=25) sets. The best model was thoroughly analyzed to understand structural requirements of the chemical pesticides for higher avian toxicity. The models developed in the current analyses may be useful to design novel less toxic pesticides.
Article Preview

Introduction

Pesticides are considered as important class of chemicals in agricultural science and research. The last few decades witnessed significant rise in the applications of pesticides for the protection of crops especially in the agriculture dependent developing countries (Akhtar, Sengupta & Chowdhury, 2009). Ideally, a pesticide is expected to elicit toxic or repellant effects on the pests (target species) which are responsible for the damage of crops and should not cause any harmful effect on other species (non-target species) of the ecosystems. However, several investigations confirmed that the currently used pesticides lack such specificity and may cause short as well as long-term toxicities towards several non-target species, including humans (Mostafalou & Abdollahi, 2013). Therefore, the ecological safety of pesticides raised huge concern over the last few decades and various regulatory agencies emphasized on the necessity of toxicity evaluations of already existing as well as novel pesticides. According to the regulatory guidelines, the safety assessment is conducted on both terrestrial and aqueous species (Benfenati, Clook, Fryday & Hart, 2007). Birds are considered as one of the most important non-target species (Zhang et al, 2015). Most of the cases oral intake of pesticides renders the birds susceptible to serious adverse effects, such as damage of immune, cardiovascular, renal, respiratory and reproductive systems (Mostafalou & Abdollahi, 2013). Currently, avian toxicity tests on some specific avian species such as northern Bobwhite quail (Colinus virginianus), Mallard duck (Anas platyrhynchos), Japanese quail, Ring-necked pheasant, House sparrow are recommended by Organization for Economic Co-operation and Development (OECD) and Environmental Protection Agency (EPA) (Basant, Gupta & Singh, 2015). The 50% lethality from oral dose (LD50) is considered as the most significant parameter for toxicity assessment.

Since in vivo testing on avian species are expensive as well as time-consuming (Jaworska, Comber, Auer & Van Leeuwan, 2003), recently different regulatory organizations suggested to limit the use of laboratory animals. The ‘3R concept’ (replacement, refinement and reduction of animals in research) encourages deployment of alternative approaches for the prediction of toxicity of chemicals in the absence of experimental data (Das & Roy, 2013). Chemometric methods especially quantitative structure toxicity relationship (QSTR) may serve as such alternative method for risk assessment of toxic chemicals. The European Union regulation ‘Registration, Evaluation, Authorization and Restriction of Chemicals (REACH)’ advocated the application of QSTR analyses to reduce the unnecessary use of laboratory animals (Jagiello, & Puzyn, 2009). In QSTR, the prediction is done on the basis of correlation between descriptors and toxicity endpoints. In order to serve the purpose of predictive assessment in QSTR, the OECD guidelines (OECD, 2007) must be followed. Previously, some investigators (Toropov & Benfenati, 2006, Mazzatorta, Cronin & Benfenati, 2006, Zhang et al., 2015, Basant et al, 2015) reported QSTR analyses on avian toxicity of pesticides. These reports showed applications of different machine learning tools to understand important structural and physicochemical characteristics responsible for higher toxicities. The current report includes Monte-Carlo based QSTR analyses (Mondal et al., 2015, Toropova et al, 2015; Veselinovic et al., 2016, Toropov et al 2011, Toropova et al, 2016) on some diverse pesticides having LD50 values determined on Bobwhite quail. The developed models were found to be predictive for external datasets and these models may be utilized for the prediction of novel pesticides.

Data Sets

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 4: 4 Issues (2019): Forthcoming, Available for Pre-Order
Volume 3: 2 Issues (2018)
Volume 2: 2 Issues (2017)
Volume 1: 2 Issues (2016)
View Complete Journal Contents Listing