Quantitative structure–activity relationship (QSAR) analysis for recently synthesized pyrazolyl-thiazolinone derivatives was studied for their EGFR and HER-2 kinase inhibitory activities. The statistically significant 2DQSAR models (r2 = 0.9086; q2 = 0.8370; F test = 49.6789; r2 se = 0.1242; q2 se = 0.1675; pred_r2 = 0.8086; pred_r2se = 0.1934 and r2 = 0.9163; q2= 0.8702; F test 54.7057; r2 se= 0.0820; q2 se=0.1020; pred_r2 = 0.8249; pred_r2se = 0.1195) were developed using molecular design suite (VLifeMDS 4.1). The study was performed with 36 compounds (data set) using sphere exclusion (SE) algorithm, random selection and manual selection methods used for the division of the data set into training and test set. Partial least square regression (PLSR) methodology with stepwise (SW) forward-backward variable selection method was used for building the QSAR models. The results of the 2D-QSAR models were further compared with 3D-QSAR models generated by kNN-MFA, (k-Nearest Neighbor Molecular Field Analysis) investigating the substitutional requirements for the favorable inhibitory activity for EGFR and HER-2 in tumor growth and providing useful information in the characterization and differentiation of their binding sites. The results derived may be useful in further designing novel EGFR and HER-2 kinase inhibitors prior to synthesis.
