Абстрактный

QSAR study of thiobenzamides using quantum mechanical descriptors

P.P.Singh, Dhruv Chandra Shukla

MLR analysis of four sets of thiobenzamides for the prediction of activity againstMIC withM. Avium,M. Tuberculosis,M. fortuitum andM. kansasii have been done using the descriptors heat of formation, molecular weight, total energy, HOMO energy, LUMO energy, absolute hardness and electronegativity. Maximum number of descriptors used in MLR analysis is 4 and total number of MLR analysis done for each set is 90. Best QSAR models developed for the four sets have the value of regression coefficient greater than 0.9 indicating the reliability of the model. Single descriptor HOMO energy has the value of regression coefficient greater than 0.5 in the first, second and fourth set of derivatives of thiobenzamides. In the third set of thiobenzamides, no single descriptor has the value of regression coefficient greater than 0.5. In this case the combination of total and LUMO energy provide the value of regression coefficient greater than 0.5.