Blinded Predictions of Standard Binding Free Energies: Lessons Learned from the SAMPL6 Challenge

<p>In the context of the SAMPL6 challenges, series of blinded predictions of standard binding free energies were made with the SOMD software for a dataset of 27 host-guest systems featuring two octa-acids hosts (<i>OA </i>and <i>TEMOA) </i>and a cucuribituril ring (<i>CB</i>8)<i> </i>host. Three different models were used, <i>ModelA </i>computes the free energy of binding based on a double annihilation technique; <i>ModelB</i> additionally takes into account long-range dispersion and standard state corrections; <i>ModelC</i> additionally introduces an empirical correction term derived from a regression analysis of SAMPL5 predictions previously made with SOMD. The performance of each model was evaluated with two different setups; <i>buffer </i>explicitly matches the ionic strength from the binding assays, whereas <i>no-buffer</i> merely neutralizes the host-guest net charge with counter-ions. <i>ModelC/no-buffer</i> shows the lowest mean-unsigned error for the overall dataset (MUE 1.29 < 1.39 < 1.50 kcal mol<sup>-1</sup>, 95% CI), while explicit modelling of the buffer improves significantly results for the CB8 host only. Correlation with experimental data ranges from excellent for the host TEMOA (R<sup>2</sup> 0.91 < 0.94 < 0.96), to poor for <i>CB8 </i>(R<sup>2</sup> 0.04 < 0.12 < 0.23). Further investigations indicate a pronounced dependence of the binding free energies on the modelled ionic strength, and variable reproducibility of the binding free energies between different simulation packages. </p>