PREDICTION OF BROAD SPECTRUM ANTIFUNGAL COMPOUNDS USING POTENCY MAPPING AND MULTIVARIATE TECHNICS

Opposed to structure-selectivity relationship the concept of structure-broad spectrum relationship has not been studied in the literature. Attempt is made for the definition of indexes measuring the broadness of spectrum of antifungal compounds. The inherent shortcoming of these indexes for QSAR uses is discussed. For the prediction of fungicides having both high activity and broad spectrum the use of the first component of principal component analysis is proposed as dependent variable in various QSAR approaches. The study is based on the calculations made on the data set of contact antifungal activities of 19 substituted N-(1-cyano-cyclohexyl)aniline derivatives against seven taxonomically different species of fungi, A successful prediction is made.