Abstract
Selective S1P(1) receptor agonists have therapeutic potential to treat a variety of immune-mediated diseases. A series of 2-imino-thiazolidin-4-one derivatives displaying potent S1P(1) receptor agonistic activity were selected to establish 3D-QSAR models using CoMFA and CoMSIA methods. Internal and external cross-validation techniques were investigated as well as some measures including region focusing, progressive scrambling, bootstraping and leave-group-out. The satisfactory CoMFA model predicted a q(2) value of 0.751 and an r(2) value of 0.973, indicating that electrostatic and steric properties play a significant role in potency. The best CoMSIA model, based on a combination of steric, electrostatic, hydrophobic and H-bond donor descriptors, predicted a q(2) value of 0.739 and an r(2) value of 0.923. The models were graphically interpreted using contour plots which gave more insight into the structural requirements for increasing the activity of a compound, providing a solid basis for future rational design of more active S1P(1) receptor agonists.