Applications of Chemometric Methods to Elucidate Physicochemical Requirements for Binding of PTP1B Inhibitors to Its Target

The quantitative structure activity relationship (QSAR) models were developed using multiple linear regression (MLR), partial least square (PLS) and feed forward neural network (FFNN) for a set of 49 PTP1B inhibitors of diabetes. The MLR,PLS and FFNN generated analogous models with good prognostic ability and all the other statistical values, such as r, r2, r2cv and F and S values, remained satisfactory. The results obtained from this study indicate the importance of dipole moment Y component, Number of H- bond and VAMP polarization (whole molecule) in determining the inhibitory activity of PTP1B inhibitor. The best artificial neural network model is a fully-connected, feed forward back propagation network with a 2-5-1 architecture. This statistics is appropriate to the further design of novel PTP1B receptor. The similarity (CARBO and HODGKIN) analysis was also done on the same series which positively support the previous results. The QSAR study reported in the present study provide important structural situation, related to anti-diabetic activity. Present study enlightens the path of determining the potent lead compounds of PTP1B antagonist.