Font Size: 

Binding free energy computation plays an important role in prioritizing compounds to be evaluated experimentally on their affinity for (off-)target proteins, yet fast and accurate binding free energy prediction remains an elusive task. In this study, we showcase the utility of two popular end-point methods, i.e., the linear interaction energy (LIE) and MM/PBSA approaches, in their ability to correlate calculated binding affinities of thieno[3,2-d]pyrimidine-6-carboxamide-derived sirtuin 1 (SIRT1) inhibitors with experimental data. By comparing LIE with standard single-trajectory setup of MM/PBSA, our study indicates that LIE enables to predict direct ('absolute') values for SIRT1 binding free energies with lower compute requirements, while the accuracy in predicting relative values for binding free energy is comparable (Pearson's r = 0.72 and 0.64 for LIE and MM/PBSA, respectively). We also find that in this particular case study, the contribution from electrostatic and polar interactions to binding free energy can be neglected to obtain LIE and MM/PBSA models of comparable predictive quality.