Affinity Prediction in Structure-Based Lead Optimization webinar was presented by Chris Neale, Principal Scientific Developer in the Affinity Group at OpenEye. Chris’ efforts are directed toward improving the accuracy and cost efficiency of OpenEye’s approach to estimating relative binding free energies of candidate ligands (alchemical non-equilibrium switching), with a special focus on enantiomer-specific binding, alchemical conversion route efficiency, and the optimization of AWS instance selection.
About this session:
Prediction of the binding affinity of a small molecule to a protein target is one of the most important contributions that computation can make to the lead optimization process.
As part of a cascade of methods for molecule profiling in our cloud-native modeling platform Orion®, including structure-based virtual screening, pose prediction, and pose stability assessment, we have implemented Non-Equilibrium Switching (NES) for relative binding affinity prediction. NES is substantially more computationally efficient than other commonly used methods for predicting relative binding affinity, such as Free Energy Perturbation (FEP) or Thermodynamic Integration (TI).
NES is also highly parallelizable, enabling very high throughput on Orion®. In extensive comparisons to other methods, NES provides useful correlation with experimental binding affinity and gives actionable guidance on classifying molecules as high or low affinity.
