Large scale benchmarking the prospective accuracy of protein-ligand free energy calculations

This presentation was given at the German Conference on Cheminformatics, November 3 2020.

Free energy simulations are rapidly becoming a key component to the drug design process to estimate the binding affinities of drug candidates to protein targets and prioritize syntheses. There have been significant efforts over the years to improve methods and parameters for free energy calculations. In particular, the Open Force Field initiative [1] develops tools and infrastructure to build high accuracy classical mechanics force fields.

Benchmarking the performance of free energy calculations on real-world systems becomes critical so that users can have an idea of prospective accuracy of the calculations. We have curated a large set of around 20 targets and 600 ligands, which we used to evaluate the Open Force Field parameters. We ran alchemical relative free energy calculations using the non-equilibrium approach based on Gromacs as implemented in pmx tool. [2]

We will present and analyze more than 1000 calculated relative free energies. The comparisons can be made with respect to other force fields such as GAFF and CGenFF. We share the learnings regarding force field parameters and other areas with the potential for improving free energy calculations. The performance of the Open Force Field parameters will be evaluated relative to experimental values and calculations with other force fields and methods.

Literature:
[1] The Open Force Field, https://openforcefield.org/, August 31, 2020.
[2] Seeliger, D.; de Groot, B. L Biophysical Journal 2010, 98, 2309–2316; Gapsys, V.; Michielssens, S.; Seeliger, D.; de Groot, B. L. J. Comput. Chem. 2015, 36, 348–354; Gapsys, V.; Pérez-Benito, L.; Aldeghi, M.; Seeliger, D.; van Vlijmen, H.; Tresadern, G.; de Groot, B. L., Chem. Sci. 2020, 11, 1140–1152.