Introduction

The aim of this tutorial is to run a Temperature Accelerated Molecular Dynamics(TAMD)/ driven-Adiabatic Free Energy Dynamics (d-AFED)  simulation using the OpenMM-UFEDMM interface. This tutorial follows the method used in the paper OpenMM-UFEDMM 2023 paper.  The basic form of Lagrangian used in this method is:

In this tutorial, we will compute the free energy surface of alanine tripeptide in vacuo as a function of four backbone dihedral angles (Φ1, Ψ1, Φ2, Ψ2) using d-AFED. It is modelled using the ff14SB all-atom AMBER force field, and the simulations will be carried out OpenMM-UFEDMM interface.

Optimal choice of the extended space parameters 

For a successful application of d-AFED and related methods, a correct choice of extended system parameters is crucial. These include the coupling constants "kappa", the mass-like parameters "mu", and the auxiliary temperature. 

1. The good choice of auxiliary CV mass is 100 times the effective mass of the corresponding CV, which can be computed as; For more details, see  J. Chem. Theory Comput. 2014, 10, 8, 2975–2986 

2. Coupling constant "kappa" can be  estimated using the following equation: 

Reweighting of d-AFED 

In d-AFED, the free energy surface F(s) at the physical temperature β can be directly computed from the distribution of the auxiliary variables s from the d-AFED simulation using the auxiliary CV temperature on the s variables.

d-AFED Simulation Files 

We need python scripts (containing all the inputs and execution instructions) and a pdb file (containing the starting structure) to run the simulations. The Python script contains all the details.

1. d-AFED Python Script 

2. Alanine tripeptide PDB  

3. Python Script for checking the parameters