Abstract

Multi linear regression analysis of carcinogencity of aromatic amines with the help of quantum chemical and energy descriptors

Author(s): P.P.Singh, S.A.Khan, Prabhat Kumar Verma

QSAR models of 73 derivatives of aniline which have carcinogenic property have been developed with the help of quantum chemical and energy descriptors such as heat of formation, global hardness, total energy, HOMO energy, LUMO energy, absolute hardness and chemical potential. The qualities of the models have been adjudged by the value of cross-validation and correlation coefficients evaluated by multi linear regression analysis. The models have been arranged in a table in descending order of predictive power. The best model has correlation coefficient about 0.90 and has been developed with the help of heat of formation, total energy, HOMO energy and chemical potential. The most widely used descriptor is total energy.