Simulating lipophilicity of organic molecules with a back-propagation neural network.

From a training set of 7200 chemicals, a back-propagation neural network (BNN) model was developed for calculating the 1-octanol/water partition coefficient (log P) of molecules containing nitrogen, oxygen, halogen, phosphorus, and/or sulfur atoms. Chemicals were described by means of autocorrelation vectors encoding hydrophobicity, molar refractivity, H-bonding acceptor ability, and H-bonding donor ability. A 35/32/1 composite network composed of four configurations was selected as the final model (root-mean-square error (RMS) = 0.37, r = 0.97) because it provided the best simulation results (RMS = 0.39, r = 0.98) on an external testing set of 519 molecules. This final model compared favorably with a recently published BNN model using variables (atoms and bonds) derived from connection matrices.

Recent trends and developments in the EU in the environmental control and management of chemicals.

In the further development and implementation of European legislation on chemicals and related products the following aspects and trends can be identified: (1) refinement and adaptation of existing legislation and guidelines to technical progress (including extension of testing and risk assessment strategies), (2) preparation and implementation of new directives and regulations, (3) increasing application and further development of screening methods and model approaches, and (4) global harmonization of hazard and risk assessment criteria and procedures in the frame of the International Forum on Chemical Safety (chapter 19 of Agenda 21). The increasing use of evaluation and screening models can be illustrated by the further development and application of specific models such as EUSES and EASE to predict effect and exposure data. On the effects assessment side, models are available or under development to predict the phototoxicity and potential endocrine disruptors (estrogen/androgen binding affinity) of industrial chemicals. Finally, the emphasis that is given worldwide to the global harmonization procedures initiated under chapter 19 of Agenda 21, agreed at the UNCED conference for the environment and sustainable development in 1992, cannot fail to influence further developments in Europe.

Molecular orbital parameters as predictors of skin sensitization potential of halo- and pseudohalobenzenes acting as SNAr electrophiles.

The electrophilic reactivity of a training set of 20 halo- and pseudohalobenzenes, 10 of which are reported skin sensitizers and 10 of which are reported nonsensitizers, has been modeled by MO-calculated indices using the AM1 and PM3 Hamiltonians. The electronic structures of parent molecules and the corresponding Meisenheimer intermediates (sigma-complexes) were evaluated. The NH2 group and the H atom were both studied as model nucleophile-derived substituents in the sigma-complexes. The LUMO energy differences between the parent compounds and their Meisenheimer complexes together with the maximum acceptor superdelocalizabilities determined over the aromatic reaction sites were found to discriminate correctly the sensitizing/reactive from nonsensitizing/unreactive compounds of the training set of 20 compounds. The predictive applicability of these MO indices was confirmed with a test set of seven further compounds for which sensitization data are reported in the literature. A statistically based discriminant analysis provides a model which predicts whether or not an SNAr electrophile will be a sensitizer and estimates the degree of confidence in the prediction.

Occupational exposure modelling with ease.

This article presents a validation exercise performed from eight practical case studies on EASE (version 2.0), a knowledge-based system allowing to estimate the workplace exposure to chemicals. Our results show that EASE represents a valuable simulation tool in occupational hygiene. However, it requires to be refined and extended to more realistic and precise situations to be easily used in practice.

The use of computer based structure-activity relationships in the risk assessment of industrial chemicals.

The concept of a two-step approach toward the assessment of toxicity endpoints for a chemical is proposed. The first step involves the selection of chemical analogues for which toxicity data is available in a noncongeneric database. The next step is the derivation of a Quantitative Structure-Activity Relationship (QSAR) for the chemical domain, predetermined by the selection rules. The software tools needed for the computer implementation of such an approach are summarized. By making use of them, we have derived aquatic toxicity QSARs, of which two are given as example. The latter pertain to chemicals that have been automatically extracted from noncongeneric databases, after defining the substructure recognition rules implied by the putative mechanism of toxicity.

Predictions for existing chemicals-a multilateral QSAR project.

Abstract Following a previous collaborative EU/EPA project focussed on QSAR predictions for a selection of new chemicals which had been notified in the EU, a similar exercise was started in 1993 on existing chemicals. In a first phase, the project addresses the High Production Volume (HPV) chemicals which are produced or imported at levels above a 1000t/year in the EU and 454t/year in the US. The relevant EU (Annex 1 of Existing Chemicals Regulation No. 793/93) and US-EPA lists contain 1036 and 2881 organic substances respectively of which HPV 749 chemicals are in common. The joint project aims at an estimation through validated QSAR models of the physical-chemical, ecotoxicity and toxicity endpoints which are included in the regulation and where experimental data will become available in IUCLID (International Unified Chemicals Information Database). Next to EC-JRC (ECB) and US-EPA, various laboratories in the EU are contributing to the project and recently, two institutes in Japan have joined in this project.

Assessment of QSARS for Predicting Fate and Effects of Chemicals in the Environment: An International European Project.

Abstract In 1993, an international project on QSAR has been started with funding from the Commission of the European Union. The first part of the project is focused on preparing an overview of existing models for the prediction of environmental parameters such as bioconcentration, sorption, degradation and ecotoxicity. Emphasis will be given to defining the limitations of the models. Since all models, including QSARs, have their limitations, it is important that these limitations are known in case QSARs are actually used and applied within the risk assessment context. The second part of the project is directed towards experimental research on new developments with emphasis on the use of multivariate techniques and quantum chemical properties. In this short paper, a general outline of the project will be given, as well as some first results. Results of experimental work within this project will be published in the proceedings of the 6th International Workshop on QSAR in Environmental Sciences and will appear in this same journal.

A General QSAR Model for Predicting the Toxicity of Organic Chemicals to Luminescent Bacteria (Microtox® test).

Abstract A large data set of Microtox® toxicity results was used to derive a general QSAR model. Chemicals were described by means of a modified autocorrelation method. The autocorrelation vectors were generated from atomic contributions encoding the hydrophobicity and molar refractivity of the molecules. A three-layer backpropagation neural network was used to design the model. The obtained results were compared with those obtained from a principal components regression analysis.

Chemometrical evaluation of multispecies-multichemical data by means of graphical techniques combined with multivariate analyses.

A new method combining graphical displays with principal component analysis has been used to evaluate published data on the toxicity of seven chemicals to 14 species (17 testing procedures) of aquatic biota. The results reflect the usefulness of simple graphical approaches for analyzing the structure of environmental data sets. Thus, the study indicates the importance of end point selection and underlines some relationships among the species and chemicals.

Nonlinear dependence of fish bioconcentration on n-octanol/water partition coefficient.

The log-log relationship between the bioconcentration tendency of organic chemicals in fish and the n -octanol/water partition coefficients breaks down for very hydrophobic compounds. The use of parabolic and bilinear models allows this problem to be overcome. The QSAR equation log BCF = 0.910 log P - 1.975 log (6.8 10(-7) P + 1) - 0.786 (n = 154; r = 0.950; s = 0.347; F = 463.51) was found to be a good predictor of bioconcentration in fish.

Estimating pesticide field half-lives from a backpropagation neural network.

The field half-lives of 110 pesticides were modelled using a backpropagation neural network (NN). The molecules were described by means of the frequency of 17 structural fragments. Before training the NN, different scaling transformations were assayed. Best results were obtained with correspondence factor analysis which also allowed a reduction of dimensionality. The training and testing sets of the NN analysis gave 95.5% and 84.6% of good classifications, respectively. Comparison with discriminant factor analysis showed that a backpropagation NN was more appropriate to model the field half-lives of pesticides.