[gmx-users] Alchemical Free energy of growing a cavity in water

Braden Kelly bkelly08 at uoguelph.ca
Fri Dec 14 07:43:09 CET 2018


I am wishing to calculate the free energy of growing a cavity in water.

It has been done by Li et al. using LAMMPS (Computational methodology for solubility prediction: Application to the sparingly soluble solutes, Lunna Li, Tim Totton, Daan Frenkel, The Journal of Chemical Physics, 146, 2017)

I cannot do other calculations I want in LAMMPS, but can in gromacs. Hence, I would like to do this in gromacs. (in case the answer was going to be: just do it in LAMMPS :) )

To do this I need to use a user defined potential energy function (strictly repulsive) U = A*exp(-rij/B + Lambda)

U is potential energy of the cavity interacting with any atom in the system. A, B given constants, rij, distance between cavity and atom, Lambda... the Lambda window, it controls the size of the cavity... Lambda varies between -10 and 5... at Lambda = -10, the cavity is essentially a point source and gone.

I have read up on making user defined tables.

However, how do I do a free energy calculation in gromacs when using a user defined function? for each window, I will manually make a table with the necessary Lambda inside it and do an individual simulation. Once this is done, gromacs only has access to the cubic fit it makes, and can't insert a neighboring lambda value. And this is necessary to do TI/BAR. For a given configuration I need to be able to sample neighboring lambda potential energies during the same simulation.

The only thing I can think of is I could manually, for each Lambda, sample phase space(just a normal NVT or NPT simulation) and save coordinates, and re-evaluate post processing the potential energy of the configurations for the user-defined potential at +1 Lambda and -1 Lambda (or all other lambdas if I was going to do MBAR). Does this sound like the best option Or even... only option? I should be able to make the standard dhdl.xvg file myself after and submit to gmx bar for evaluation?

The next part of the problem is to grow a molecule inside the fully grown cavity. This molecule will be "normal" LJ and Coul interaction, but will need to be anchored to the COM of the cavity as the cavity, fully grown, explores phase space... to keep the molecule in the cavity I should use a restraint... is the pull code the thing to use?


Braden Kelly

PhD. Candidate, E.I.T

University of Guelph

Biophysics Interdisciplinary Group (BIG)

"I feel more like I do now than I did when I first got here."


More information about the gromacs.org_gmx-users mailing list