QSAR quantitative structure-activity neural networks

A challenging task in material science as well as in pharmaceutical research is to custom tailor a compound s properties. George S. Hammond stated that the most fundamental and lasting objective of synthesis is not production of new compounds, but production of properties (Norris Award Lecture, 1968). The molecular structure of an organic or inorganic compound determines its properties. Nevertheless, methods for the direct prediction of a compound s properties based on its molecular structure are usually not available (Figure 8-1). Therefore, the establishment of Quantitative Structure-Property Relationships (QSPRs) and Quantitative Structure-Activity Relationships (QSARs) uses an indirect approach in order to tackle this problem. In the first step, numerical descriptors encoding information about the molecular structure are calculated for a set of compounds. Secondly, statistical and artificial neural network models are used to predict the property or activity of interest based on these descriptors or a suitable subset. 

Methods other than thermodynamic cycles are often used to calculate acid dissociation constants. Previous publications implement the theoretical relationship between pKa and structural property [6], bond valence methods and bond lengths [33], pKa correlations with highest occupied molecular orbital (HOMO) energies and frontier molecular orbitals [34], and artificial neural networks [35] to predict pKa values. In addition much work has been done using physical properties as quantitative structure-activity relationship (QSAR) descriptors, and regression equations with such descriptors to yield accurate pKa values for specific classes of molecules [36-47]. The correlation of pKas to various molecular properties, however, is often restricted to specific classes of compounds, and it is... 

In terms of practical application, expert systems overlap with systems for deriving and applying quantitative structure-activity relationship (QSAR) models or equations, and with systems using artificial neural networks (ANN) or genetic algorithms. The expert systems described in this chapter are characterized by their use of a generalized store of knowledge. 

All developments of quantitative structure activity relationships (QSARs)/ quantitative structure-property relationships (QSPRs)/QSDRs go through similar steps (1) collection of a database of measured values for model development and validation/evaluation, (2) selection of chemical descriptors (can include connection indices, atom, bond, or functional groups, molecular orbital calculations), (3) development of the model (develop a correlation between the chemical descriptors and the activity/property/degradation values) using a variety of statistical approaches (linear and non-linear regression, neural networks, partial least squares (PLS), etc. [9]), and (4) validate/evaluate the model for predictability (usually try to use a separate set of chemicals other than the ones used to train the model external validation) [10]. 

Barratt, M. D. Quantitative structure-activity relationships (QSARs) for skin corrosivity of organic acids, bases and phenols principal components and neural network analysis of extended datasets. Toxicol, in Vitro 1996,10, 85-94. 

Also in chemistry artificial neural networks have found wide use. They have been used to fit spectroscopic data, to investigate quantitative structure-activity relationships (QSAR), to predict deposition rates in chemical vapor deposition, to predict binding sites of biomolecules, to derive pair potentials from diffraction data on liquids, " to solve the Schrodinger equation for simple model potentials like the harmonic oscillator, to estimate the fitness function in genetic algorithm optimizations, in experimental data analysis, to predict the secondary structure of proteins, to predict atomic energy levels, " and to solve classification problems from clinical chemistry, in particular the differentiation between diseases on the basis of characteristic laboratory data. ... 

Before reviewing the existing applications of neural networks to combinatorial discovery, we offer brief descriptions of the key concepts in quantitative structure-activity relationships (QSAR), neural networks, and virtual high-throughput screening (VHTS). There are a substantial number of reviews of the applications of QSAR to chemistry and dmg design (2-13) and of applications of neural networks to chemistry (13-29), to which the reader is referred for more detailed discussions of these topics. 

Sutter and co-workers reported a method for automated descriptor selection for quantitative structure-activity relationships using generalized simulated annealing (36,132). The cost function used to evaluate the effectiveness of the deseriptors was based on a neural network. The result is an automated descriptor selection algorithm that is an optimization inside of an optimization. Application of the method to QSAR shows that effective descriptor subsets are found, and they support models that are as good or better than those obtained using traditional linear regression methods. 

Balaban, A. T. and Basak, S. C. (2000) Trends and possibilities for future developments of topological indices. Abstr. Pap. Am. Chem. Soc. 220th, COMP-048. Sutter, J. M. and Jurs, P. C. (1995) Selection of molecular descriptors for quantitative structure-activity relationships. Data//anrfZ. Sci. Technol. 15, 111-132. Winkler, D. A. and Burden, F. R. (2000) Robust QSAR models from novel descriptors and Bayesian regularized neural networks. Mol. Simul. 24,243-258. Maddalena, D. J. (1998) Applications of soft computing in drug design. Expert Opin. Ther. Pat. 8, 249-258. 

Tetko, I. V., Aksenova, T. I., Volkovich, V. V., Kasheva, T. N., Filipov, D. V., Welsh, W. J., et al. (2000) Polynomial neural network for linear and non-Unear model selection in quantitative-structure activity relationship studies on the Internet. SAR QSAR Environ. Res. 11, 263-280. 

---