Development of QSAR Models of Triazine Derivatives Using Topological Descriptors

Kalpana Singh

Department of Chemistry,
Shri J.N.P.G. College, Lucknow (UP), INDIA.


QSAR Models of 25 derivatives of triazine have been developed using topological descriptors. The topological descriptors have a combination of connectivity index, shape index, molar refractivity, molecular weight, log P and conformation minimum energy. Best QSAR has been developed using connectivity index, shape index, molar refractivity, and conformation minimum energy with the regression coefficient 0.966117 and cross-validation 0.953201. Multilinear regression (MLR) equation is said to describe a good QSAR model if the value of correlation or regression coefficient and cross validation coefficient is greater than 0.2. These values indicate very good prediction ability of QSAR model. Topological descriptors which alone can produce good QSAR models are conformation minimum energy, shape index and molar refractivity.

Keywords :QSAR, Multilinear regression, connectivity index, PM3, conformation minimum energy.