Publication date: May 03, 2020
Alchemistic simulations are versatile tools for prediction of relative free energy differences. Accuracy of these methods depends critically on sampling of orthogonal (non-Alchemistic) degrees of freedom. Here we apply flying Gaussian method to accelerate such orthogonal degree of freedom – peptide bond cis/trans iso-merisation. The approach is demonstrated on prediction of pKa value of N-acetylproline. Isomerization of the amide bond was accelerated in this simulation by multiple orders of magnitude. Alchemistic free energy was obtained by reweighting. We also demonstrate that the accuracy of biased Alchemistic simulations can be significantly improved by a simple redefinition of the thermodynamic cycle using a flattening. Such redefinition can be applied a posteriori to improve the accuracy of biased Alchemistic simulations.
No Explore or Discover sections associated with this archive record.
|4.0 MiB||Input files (input structures, topology, Gromacs and Plumed input files, analysis scripts) for simulations used to demonstrate functionality of combined thermodynamic integration and flying Gaussian method. The method is demonstrated on N-acetylproline. Tested on OpenMPI 4.0.0, Gromacs 2018.5 and Plumed 2.5.0. Submission scripts are provided.|
|7.0 KiB||README.txt file with descriptions.|