Quantitative structure–activity relationship modeling of s-triazines and 2-arylpyrimidines as selective PDE4B inhibitors

Introduction: A QSAR analysis was performed on a series of 2-arylpyrimidine and s-triazine derivatives as selective PDE4B inhibitors. Primary objective of the study was to develop predictive QSAR models for s-triazines and 2-arylpyrimidines as selective PDE4B inhibitors and to identify structural features which are responsible for the PDE4B selectivity. Materials and Methods: A data set comprising 62 compounds as PDE4B inhibitors was used for development of first QSAR model while data set of 57 compounds as PDE4D inhibitors was used to develop another QSAR model. Data set was divided into training (80%) and test (20%) set by using K-mean clustering method. CDK and chemaxon descriptors were obtained for all compounds. QSAR model was built using multiple linear regression (MLR) technique. Squared cross-validation leave one out (LOO) coefficient (R2cv) for internal validation was calculated whereas external validation was performed by predicting the activity of test set using QSAR model developed. Conclusion: The results suggest that ATSm4, Wlambda3.unity, C1SP1, RNCS, TPSA, asa_ASA_P_pH_7.4 and maximalprojectionradius are important in determining the PDE4B inhibition, while BCUT-1l, WNSA-3, nAtomP, TPSA and C1SP3 are vital structural features in determining the PDE4D inhibition. TPSA and C1SP3 are negatively correlated with the PDE4D inhibition.