Artificial Neural Network Modeling of the Environmental Fate and Ecotoxicity of Chemicals. In: Devillers J, editor. Ecotoxicology Modeling. Boston: Springer US; 2009. pp. 1-28. [DOI: 10.1007/978-1-4419-0197-2_1]

QSAR modelling of synergists to increase the efficacy of deltamethrin against pyrethroid-resistant Aedes aegypti mosquitoes$

Space spraying of deltamethrin allows the control of adult Aedes (Stegomyia) aegypti mosquitoes. Unfortunately, many vector control programs are threatened by the development of resistances that decrease the efficacy of this adulticide. Faced with this situation, we can either try to use another insecticide presenting a different mechanism of action or find a strategy that brings back the efficacy of the insecticide at a satisfying level to pursue its use in vector control. Restoration of the efficacy of an insecticide can be obtained by means of a synergist. In this context, QSAR modelling was used to find synergists to combine with deltamethrin for increasing its efficacy against resistant strains of Ae. aegypti. Seventy-four structurally diverse chemicals with their 24-hour LD₅₀ values, obtained under the same experimental conditions on Ae. aegypti females, were used. Molecules were described by means of autocorrelation vectors encoding lipophilicity, molar refractivity, H-bonding acceptor and donor ability. A three-layer perceptron (TLP) was employed as statistical tool. The performances of the models were evaluated through the analysis of the prediction results obtained on the different training and test sets (80%/20%) as well as from an out-sample test set. A 6/4/1 TLP computed with the Broyden-Fletcher-Goldfarb-Shanno second-order training algorithm led to the best prediction results. The convergence was obtained in 132 cycles. The sum of squares was used as error function. The hidden and output activation functions were tanh and exponential, respectively. Various chemical structures were identified as potential synergists and searched for their commercial availability. Molecules of interest were tested in vivo on Ae. aegypti by using the susceptible reference Bora Bora strain and two resistant strains from Martinique island. This led to the identification of the PSM-05 molecule that shows interesting synergistic activity.

Use of artificial neural networks in the QSAR prediction of physicochemical properties and toxicities for REACH legislation

With the introduction of the REACH legislation in the European Union, there is a requirement for property and toxicity data on chemicals produced in or imported into the EU at levels of 1 tonne/year or more. This has meant an increase in the in silico prediction of such data. One of the chief predictive approaches is QSAR (quantitative structure-activity relationships), which is widely used in many fields. A QSAR approach that is increasingly being used is that of artificial neural networks (ANNs), and this chapter discusses its application to the range of physicochemical properties and toxicities required by REACH. ANNs generally outperform the main QSAR approach of multiple linear regression (MLR), although other approaches such as support vector machines sometimes outperform ANNs. Most ANN QSARs reported to date comply with only two of the five OECD Guidelines for the Validation of (Q)SARs.