Purpose. Radial basis function artificial neural networks and theoretical descriptors were used to develop a quantitative structure– pharmacokinetic relationship for structurally diverse drug compounds.
Methods. Human bioavailability values were taken from the literature and descriptors were generated from the drug structures. All models were trained with 137 compounds and tested with a further 15, after which they were evaluated for predictive ability with an additional 15 compounds.
Results. The final model possessed a 10-31-1 topology and training and testing correlation coefficients were 0.736 and 0.897, respectively. Predictions for independent compounds agreed well with experimental literature values, especially for compounds that were well absorbed and/or had high observed bioavailability. Important theoretical descriptors included solubility parameters, electronic descriptors, and topological indices.
Conclusions. Useful information regarding drug bioavailability was gained from drug structure alone, reducing the need for experimental methods in drug development.

The action of estrogens is mediated through the estrogen receptor alpha (ERα) and the more recently discovered estrogen receptor beta (ERβ). These estrogen receptor (ER) subtypes have distinct functions and differential tissue distribution patterns. Tissue- or cell-specific estrogenic activity of receptor ligands have become targets of drug research due to the potential to affect and control physiological and disease states such as breast and endometrial carcinoma, osteoporosis, and menopause. Receptor-ligand activity can be achieved in different ways such as by selective binding or selective modulation. These, in turn, are governed by the intermolecular interactions between estrogen receptors and their ligands.
The estrogen receptor ligand binding pocket has a degree of flexibility enabling binding of endogenous and synthetically-derived steroids, as well as non-steroidal molecules. Ligand fit is dependent upon aspects of size, polarity, and specific subsitution on ring and sidechain structures. Selectivity of a ligand for the estrogen receptor subtypes can be explained on the basis of differences in ligand-binding affinity, ligand potency, or ligand efficacy. In addition, molecular characteristics can lead to selective antagonism by ligands as well as antiestrogen character. Determinants of selectivity and antagonism have been elucidated using x-ray crystallography revealing various intermolecular and steric features of importance.
The present review will examine aspects of estrogenic binding including nonselective binding, and ERα/ERβ selectivity. Various chemical classes are critically examined including endogenous compounds, phytoestrogens, and other classes of interest to drug discovery and pharmaceutical product development.

Phytoestrogens are a diverse group of plant-derived compounds that structurally or functionally mimic mammalian estrogens and show potential benefits for human health. An increase in phytoestrogen research over the past few decades has demonstrated the biological complexity of phytoestrogens, which belong to several different chemical classes and act through diverse mechanisms. Identification of the estrogen receptor beta (ER beta) and research into various ligand classes has enabled elucidation of molecular aspects important in selective ER binding. This article explores the structural characteristics and significance of functional groups as they relate to phytoestrogen selectivity for ER binding.

The goal of quantitative structure–pharmacokinetic relationship analyses is to develop useful models that can predict one or more pharmacokinetic properties of a particular compound. In the present study, a multiple-output artificial neural network model was constructed to predict human half-life, renal and total body clearance, fraction excreted in urine, volume of distribution, and fraction bound to plasma proteins for a series of cephalosporins. Descriptors generated solely from drug structure were used as inputs for the model, and the six pharmacokinetic parameters were simultaneously predicted as outputs. The final 10 descriptor model contained sufficient information for successful predictions using both internal and external test compounds. Descriptors were found to contribute to individual pharmacokinetic parameters to differing extents, such that descriptor importance was independent of the relationships between pharmacokinetic parameters. This technique provides the advantage of simultaneous prediction of multiple parameters using information obtained by nonexperimental means, with the potential for use during the early stages of drug development.

Evidence suggests that environmental exposure to estrogen-like compounds can cause adverse effects in humans and wildlife. The Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC) has advised screening of 87,000 compounds in the interest of human safety. This may best be accomplished by pre-screening using quantitative structure-activity relationship (QSAR) modelling. The present study aimed to develop in silico QSARs based on natural, semi-synthetic, synthetic, and phytoestrogens, to predict the potential estrogenic toxicity of pesticides. A diverse set of 170 compounds including steroidal-, synthetic- and phytoestrogens, as well as pesticides was used to construct the QSAR models using artificial neural networks (ANNs). Mean correlation coefficients between experimentally measured and predicted binding affinities were all greater than 0.7 and models had few false negative results, an important consideration for screening tools. This study demonstrated the utility of ANNs as QSAR models for pre-screening of potential endocrine disruptors.

Oral dosing is the most common method of drug administration, and final plasma concentrations of the drug depend upon its bioavailability. In the current study, a quantitative structure-pharmacokinetic relationship (QSPR) was developed for a diverse range of compounds to allow prediction of drug bioavailability. Bioavailability data for 169 compounds was taken from the literature, and from the molecular structures 94 theoretical descriptors were generated. Stepwise regression was employed to develop a regression equation based on 159 training compounds, and predictive ability was tested on 10 compounds reserved for that purpose. The final regression equation included eight descriptors that represented electronic, steric, hydrophobic and constituent parameters of the drug molecules, all of which could be related to solubility and partitioning properties. Predicted bioavailability for the training set agreed more closely for drugs exhibiting mid-range literature bioavailability values. A correlation of 0.72 was achieved for test set bioavailability predictions when compared with literature values. The structure-pharmacokinetic relationship developed in the current study highlighted solubility and partitioning characteristics that may be useful in designing drugs with appropriate bioavailability.

Since the majority of lead compounds identified for drug clinical trials fail to reach the market due to poor efficacy in humans or poor pharmacokinetics (PKs), the prediction of PK properties in humans plays an important role in selection of potential drug candidates. The aim of the present study was to develop novel models for the prediction of separate PK parameters for a diverse set of drugs. Prediction would be based on the retention of each drug using micellar liquid chromatography (MLC) and selected theoretically-derived descriptors. Retention time, half life (t1/2), and volume of distribution (Vd) for each of the 26 training drugs were extracted from literature while molecular descriptors were generated using Molecular Modeling Pro. A total of 35 molecular descriptors describing molecular size, shape and solubility were calculated from the 3D molecular structure of each compound. Artificial neural network (ANN) modeling was used to correlate the calculated descriptors and retention time with half life and volume of distribution. A sensitivity analysis procedure was used to refine the models. The final predictive models showed significant correlations with literature values of t1/2 and Vd: 0.854 and 0.855 respectively for the internal testing data and 0.720 and 0.827 respectively for the external validation set of compounds. Absolute predicted values were in good agreement with literature values. Analysis of descriptors in the optimum models revealed a large degree of overlap. Solubility characteristics, hydrogen bonding, and molecular size and shape were shown to play important roles in determining drug t1/2 and Vd. The reciprocal of retention time was also included in both optimum models attesting to the significance of this particular physicochemical parameter and the complexity of the models developed. This novel combination of theoretical and experimental data for pharmacokinetic modeling may lead to further progress in drug development.

Antipsychotic medications have a diverse pharmacology with affinity for serotonergic, dopaminergic, adrenergic, histaminergic and cholinergic receptors. Their clinical use now also includes the treatment of mood disorders, thought to be mediated by serotonergic receptor activity. The aim of our study was to characterise the molecular properties of antipsychotic agents, and to develop a model that would indicate molecular specificity for the dopamine (D2) receptor and the serotonin (5-HT) transporter. Back-propagation artificial neural networks (ANNs) were trained on a dataset of 47 ligands categorically assigned antidepressant or antipsychotic utility. The structure of each compound was encoded with 63 calculated molecular descriptors. ANN parameters including hidden neurons and input descriptors were optimised based on sensitivity analyses, with optimum models containing between four and 14 descriptors. Predicted binding preferences were in excellent agreement with clinical antipsychotic or antidepressant utility. Validated models were further tested by use of an external prediction set of five drugs with unknown mechanism of action. The SAR models developed revealed the importance of simple molecular characteristics for differential binding to the D2 receptor and the 5-HT transporter. These included molecular size and shape, solubility parameters, hydrogen donating potential, electrostatic parameters, stereochemistry and presence of nitrogen. The developed models and techniques employed are expected to be useful in the rational design of future therapeutic agents.