QSAR of anti tuberculosis drugs of INH type using graphical invariants

Abstract

Quantitative structure–activity relationship studies have been performed on a series of analogues of tuberculostatic drugs of isonicotinic acid hydrazide type using multi parameter regression analysis. Theoretical molecular descriptors like Wiener index, molecular complexity indices and inter molecular similarity have been introduced in the physicochemical property based models and the superiority of the models over property–property correlation is clearly established. Theoretical molecular descriptors can be calculated directly from molecular structure, so its use in QSAR models is of immense utility compared to experimentally determined physicochemical parameters that are not always available for a particular structure. The influence of non-empirical parameters on the activity of INH and its derivatives are discussed in details. The justification for the choice of molecular descriptors based on ridge regression has been made and linearity of the models discussed.