A simple electrotopological index for quantitative structure-activity relationship correlation of physical properties with biomolecular activities

Abstract

Quantitative structure-activity relationship (QSAR) studies constitute a process by which the physicochemical properties of a set of chemical structures are quantitatively correlated with a measurable, such as the concentration of a substance required to give a certain therapeutic drug response. 2D-QSAR studies start with 10–20 analogues, ranging from biologically active to inactive; each analogue, regardless of bioactivity, is described by a series of descriptors. To further broaden the practical utility of these simple descriptors we have sought to identify hybrid indices which are trivial to calculate but which capture data from at least two categories of descriptors. An electrotopological descriptor, termed ET$_\mathrm{Z}$, which combines electronic information and molecular topology, has been devised and validated against a set of 25 anticonvulsant hydantoin molecules. This ET$_\mathrm{Z}$ is based solely on atomic connectivity information obtained from the graph without explicit input from molecular geometry.