Multilayer Perceptron Model for Predicting Acute Toxicity of Fungicides on Rats

Multilayer Perceptron Model for Predicting Acute Toxicity of Fungicides on Rats

Mabrouk Hamadache (Faculté des sciences et de la technologie, Département du génie des procédés et environnement, Laboratoire des Biomatériaux et Phénomènes de Transport (LBMPT), Université de Médéa, Médéa, Algeria), Abdeltif Amrane (Ecole Nationale Supérieure de Chimie de Rennes, Université de Rennes 1, Rennes, France), Salah Hanini (Faculté des sciences et de la technologie, Département du génie des procédés et environnement, Laboratoire des Biomatériaux et Phénomènes de Transport (LBMPT), Université de Médéa, Médéa, Algeria) and Othmane Benkortbi (Faculté des sciences et de la technologie, Département du génie des procédés et environnement, Laboratoire des Biomatériaux et Phénomènes de Transport (LBMPT), Université de Médéa, Médéa, Algeria)
DOI: 10.4018/IJQSPR.2018010106
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Quantitative Structure Activity Relationship (QSAR) models are expected to play an important role in the risk assessment of chemicals on humans and the environment. In this study, a QSAR model based on 10 molecular descriptors to predict acute oral toxicity of 91 fungicides to rats was developed and validated. Good results (PRESS/SSY = 0.085 and VIF < 5) were obtained, showing the validation of descriptors in the obtained model. The best results were obtained with a 10/11/1 Artificial Neural Network model trained with the Levenberg-Marquardt algorithm. The prediction accuracy for the external validation set was estimated by the Q2ext which was equal to 0.960. Accordingly, the model developed in this study provided excellent predictions and can be used to predict the acute oral toxicity of fungicides, particularly for those that have not been tested as well as new fungicides.
Article Preview

Introduction

Large amounts of pesticides are applied worldwide. In this regard, pesticides are of paramount importance since they allow controlling insect or fungal infestations or growth of weeds, either to anticipate long-lasting problems (Regueiro et al., 2015). Among them, fungicides are widely used in grain crops such as wheat, rice, beans, and peanuts as well as food protection and for the production of crops of suitable quality and quantity (Costaa et al., 2015). More than 3600 fungicides have been globally registered (Li et al., 2016).

Numerous monitoring studies throughout the world have demonstrated the potential of fungicides to contaminate environment. The consequences of this pollution are the widespread presence of residues in air, surface and ground waters, soil and foodstuffs (Fernandez et al., 2015; Kaushik et al., 2009; Papadakis et al., 2015; Nagel et al., 2014). In addition, they also pose a threat to the environment, humans, animals and other organisms.

Fungicides may pose risk to human health and wildlife. Several studies have showed a connection between some chronic diseases and currently used fungicides. Recent studies suggest that exposure to certain agrochemicals, including fungicides such maneb and mancozeb, may be involved in the development of neurodegenerative disorders such as Parkinson's disease (Regueiro et al., 2015). In humans, it is suggested that the exposure to some pesticides, including fungicides, decreases sperm concentration and deregulates menstrual cycles (Costaa et al., 2015). Particular fungicides have been linked with oxidative stress and mitochondrial dysfunction (Nagel et al., 2014).

Despite the likely ecological risks, fungicides have received relatively little attention in comparison with other types of pesticides, such as insecticides and herbicides (Wightwick et al., 2012). More attention should be therefore paid to risk of fungicides to human health.

It is evident that risk assessment for fungicides can provide a precaution against the corresponding pollution. One of the procedures currently used for human and environmental risk assessment is the determination of the acute toxicity of fungicides (Lagunin et al., 2007). Unfortunately, experimental determination of the toxicity is time-consuming, requires a high expense and poses an ethical problem (demands to reduce or abolish the use of animals). The use of in silico predictive methods, based on computer tools, offers a rapid, cost-effective and ethical alternative to testing toxicity of chemical substances in animals (Sullivan et al., 2014). These methods include the Quantitative Structure–Activity Relationship (QSARs) models. The objective of QSAR is to develop mathematical equations which can be able to establish the relationships between biological activity and descriptors derived solely from the molecular structure of the compound. Once a correlation is established and validated, it may be applicable to predicting the biological activity of new compounds that have not been synthesized or found. This correlation can also provide an overview of the mechanism of this activity (Can et al., 2013). To establish a QSAR model, three elements are necessary. The first relates to the biological activity (eg toxicity) measured for a set of molecules. The second concerns the descriptors. Finally, the third must be a statistical learning method that is used to connect the first two elements.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 3: 2 Issues (2018): 1 Released, 1 Forthcoming
Volume 2: 2 Issues (2017)
Volume 1: 2 Issues (2016)
View Complete Journal Contents Listing