Screening of Prominent Physicochemical Parame...

Screening of Prominent Physicochemical Parameters for Structural Requirement of Potent Toxoplasma gondii DHFR Inhibitors

Author Name : Ruchi Mishra

ABSTRACT

Molecular modeling studies was performed on 29 pyrimidine and quinazoline analogues with anti-toxoplasma activity using multiple linear regression (MLR), partial least squares (PLS) and feed forward neural network (FFNN) techniques. The MLR and PLS generated comparable models with good predictive ability and all other statistical values, r, r², r²_cv, r² (test set), F and S values, were 0.95, 0.89, 0.89, 0.81 and 90.14, 0.26 respectively, for MLR and r², r²_cv, r² (test set) and statistical significance value were 0.89, 0.89, 0.81 and 0.99 respectively, for PLS, were satisfactory. The model developed from feed forward neural network (FFNN) technique also showed good correlation value of r²= 0.90 and 0.81 respectively for training and test set. The analysis helped to ascertain the role of kier chi V3 (cluster) index and number of H- bond donors for whole molecule in determining the Toxoplsma gondii DHFR inhibitory activity of the compounds.

Keywords: Quantitative structure activity relationship, pyrimidine and quinazoline analogues, Toxoplsma gondii DHFR inhibitory activity.