Docking and QSAR Studies for Inhibitors of Thymidylate Synthase

A series of novel antifolate inhibitors having naphthalene core, substituted quinazoline, indole, pyrrolo– pyrimidine, pyrido–pyrimidine, and pteridine were designed using computational technique. These molecules were compared with the known classical and non–classical antifolate inhibitors of the thymidylate synthase by performing docking studies and by computing their ADME properties. The designed molecules showed good binding affinity towards the protein compared to the several thymidylate synthase inhibitors. The biological activities for these inhibitors were predicted withy a model equation generated by regression analysis between biological activity (pKi) of known inhibitors and there E–model which is a specific combination of Glide score, Coulombic and van der Waals interactions. The MLR QSAR analysis was carried out on 20 analogues used as training set, and 8 analogues used as test set. This study gave a reasonably good predictive model with R 2 = 0.957 and R 2 LOO = 0.871 (leave–one–out method). The cross validation on the test set gave R 2 cv = 0.587 and RMS = 0.493.