Events
Automated determination of optimal λ schedules for free energy calculations
Date and duration
Date: April 27th 2022
Time: 2pm GMT / 9am EST
Duration: 45 minutes
Format: Webinar
Abstract
Relative free energy calculations are being increasingly used in drug discovery due to the combination of improved algorithms, faster hardware and better usability. These calculations generally involve the use of alchemical intermediates, where one ligand is transformed into another one in a non-physical way. The choice of exactly how to do this transformation can have a strong effect on the stability and accuracy of the free energy calculation. Traditionally the transformation coordinate λ is either varied linearly between the endpoints, or a sigmoidal function is used to provide extra resolution near the endpoints. However, the choice of which λ values to use is not always clear: if too many values are used then computational resources are wasted, while if too large a gap is left between adjacent λ values accuracy can suffer.
In this webinar we present a new method for automatically determining optimal λ schedules for a given transformation. A short pre-calculation is used to obtain an estimate of the phase space overlap matrix. Analysis of this matrix can reveal regions where the λ values are too far apart for good convergence, in which case an iterative procedure can be applied to introduce additional λ values and repeat the calculation. The output is a set of optimised λ values which ensure a good phase space overlap between all adjacent λ values while minimising the total number of λ windows that need to be simulated. Validation on standard data sets indicates that the simulation time can be reduced by 30% with no loss of accuracy.
About the presenter
Dr Mark Mackey, Chief Scientific Officer
Mark was one of the founders of Cresset when the company was formed in 2002. He became CSO in 2013 and heads up our scientific team. He has made fundamental contributions to our underlying scientific approach and to the development of the software. He works hard to extend the scientific base of the company in a way that is consistent with our reputation of rigorous scientific excellence. His team’s current research includes new ways of applying the XED force fields to proteins and improving the handling of water molecules. Mark’s previous experience includes roles at Napp Pharmaceuticals and Merck, Sharpe and Dohme.
His PhD is in Chemistry from the University of Cambridge and he is a Fellow of the Royal Society of Chemistry.
Register
Registration for this event is now closed.