Prediction of the genotoxicity of aromatic and heteroaromatic amines using electrotopological state indices.

In the past decade, electrotopological state (E-state) indices have come into their own as useful descriptors for correlating a variety of physicochemical and biological properties of chemical compounds. Genotoxicity and mutagenicity, however, appear not to have been previously considered. In the present study, the genotoxicity of a set of 95 aromatic and heteroaromatic amines, which has been modeled previously using several sets of parameters, is modeled using E-state indices, both with and without principal components analysis. Parallels are drawn between E-state indices that were important in these models and other types of descriptors found significant in previous studies, thus, shedding light on connections to the molecular mechanism of activity. The best result had a correlation coefficient r = 0.876 and a standard error s< or = 1 log unit. These values are comparable to those in previously published models that were based on topological/geometric or on physicochemical parameters. They are not as good as those for a model based on descriptors derived from extensive quantum mechanical analysis, but E-state indices are much easier to compute.

Polynomial expressions for the hyper-Wiener index of extended hydrocarbon networks.

General expressions are derived for the hyper-Wiener index (WW) for several series of hydrocarbons, both benzenoid and non-benzenoid, including some two-dimensional networks. Indeed, such expressions were found for all but one series studied. Methods are also described for eliminating the register overflow problem that besets computer-based approaches to calculating WW for large structures by means of the distance matrix.

The permanental polynomial.

This study identifies properties and uses of the permanental polynomial of adjacency matrixes of unweighted chemical graphs. Coefficients and zeroes of the polynomial for several representative structures are provided, and their properties are discussed. A computer program for calculating the permanental polynomial from the adjacency matrix is also described.

Permanental polynomials of the smaller fullerenes.

Using a general computer code developed previously, permanental polynomials were computed for all fullerenes C< or =36. Mathematical properties of the coefficients and zeroes were investigated. For a given isomer series of constant n, the n/2 independent zeroes appear to consist of a set of 10 that are nearly constant within the series and a set of n/2-10 that differ greatly with structure.

Estimation of octanol-water partition coefficients from a molecular similarity parameter.

For 1966 compounds for which experimental values of the octanol-water partition coefficient (as logKow) were known, the value of an intermolecular similarity parameter, S, was computed for all 3,863,190 possible pairs. The S parameter differs from those most often used to predict molecular properties in that S depends on the difference between whole structures rather than on the difference between two numbers derived from the structures individually. For each compound, the largest S values were used to estimate logKow, and the statistical quality of the results was compared with results for the same 1966 compounds obtained by fragment-constant methods.

Comparison of structure-activity relationships derived from two methods for estimating octanol-water partition coefficients.

This study examines several established linear relationships between 96-hour acute toxicity to fish (as log LC50) and octanol-water partition coefficient (as log Kow) with regard to the effect of using a different method of calculating log Kow. For all seven classes of compounds examined here, the linear equation parameters, viz., slope, intercept, r2, and standard error of the estimated log LC50, were about the same for the two methods. As expected, the accuracies of predictions for individual compounds were often different for the two methods, but seldom remarkably so.

Correlation of physicochemical properties of alkylphenols with their graph-theoretical epsilon parameter.

For a series of 40 alkylphenols, the graph-theoretical parameter epsilon correlates very closely with physicochemical parameters of interest in predicting environmental toxicity and fate, and especially with logKOW (r = 0.998). Possible uses of this correlation are discussed, along with needs for extending the epsilon parameter to other types of molecules.

Correlation of graph-theoretical parameters with biological activity.

Osmialowski and Kaliszan calculated graph-theoretical indices for substituted isonicotinic hydrazides and used simple and multiple regression to search (unsuccessfully) for correlations with biological activity. The present paper describes successful searches for correlation in the same data set using principal component analysis (PCA) with multivariate outlier testing and also using stepwise multiple regression. Following PCA, correlation with biological activity always appeared in the second principal component, not the first, that is, after projection of the data points into the (n - 1)-space orthogonal to the first principal component axis. In that space, the principal component score was a more accurate predictor of biological activity than were equations provided by multiple regression or stepwise multiple regression using the underlying variables. A multivariate outlier test identified one observation as discordant, and removing that observation improved prediction further.