Interpretation of the binding affinities of PTP1B inhibitors with the MM-GB/SA method and the X-score scoring function

We have studied the binding affinities of a set of 45 small-molecule inhibitors to protein tyrosine phosphatase 1B (PTP1B) through computational approaches. All of these compounds share a common oxalylamino benzoic acid (OBA) moiety. The complex structure of each compound was modeled by using the GOLD program plus the ASP scoring function. Each complex structure was then subjected to a molecular dynamics (MD) simulation of 2 ns long by using the AMBER program. Based on the configurational ensembles retrieved from MD trajectories, both MM-GB/SA and MM-PB/SA were employed to compute the binding free energies of all 45 PTP1B inhibitors. The correlation coefficient between the MM-GB/SA results and experimental binding data was 0.87 and the standard deviation was 0.60 kcal/mol. The performance of MM-PB/SA was slightly inferior to that of MM-GB/SA. Several aspects of the MM-GB(PB)/SA method were explored in our study to obtain optimized results. The X-Score scoring function was found to produce equally good results as MM-GB/SA on both the complex structures prepared by molecular docking and the configurational ensembles obtained through lengthy MD simulations. The structure-activity relationship of this set of compounds is also discussed based on the computed results. The computational approaches validated in our study are hopefully applicable to the study of other classes of PTP1B inhibitors.