Pulling a ligand out of the active site of a protein for PMF analysis

GROMACS version: 2019.4
GROMACS modification: No

I am trying to pull a docked ligand out of the active site of my protein in a specific direction using the COM module of Gromacs and then perform PMF analysis. I have never done this before and I have a few questions that I hope someone can help me with.

  1. When choosing the reaction coordinate, it is specified by 2 different groups, pull_group1 and pull_group2. For my particular example, pull_group1 should be the ligand, while pull_group2 can be, for example, a residue in the active site of the protein? Or should the second group be the entire protein?

  2. When choosing pull_group1, is it best to choose the entire ligand, or just a single atom of the ligand?

  3. I need to pull the ligand in a specific direction. Can this be defined by first setting the pull_coord1_geometry setting to “direction”, and then specifying the coordinates of the vector along which the ligand is pulled in the pull_coord1_vec option?

Best wishes,


It depends on your reaction coordinate. The choice of reaction coordinate is difficult to generalize, it depends from the system and question you want to answer. Exploratory studies are often needed to provide the insight to choose the reaction coordinate. I suggest also to have an exploratory study on which is the most suitable pull_coord1_geometry for your problem and system.

Note that pull-coord-geometry = direction can cause unwanted torque forces in the system, unless you pull against a reference group with (nearly) fixed orientation, e.g. a membrane protein embedded in a membrane along x/y while pulling along z. If your reference group does not have a fixed orientation, you should probably other geometry option.

(see more here Non-equilibrium pulling — GROMACS 2021.3 documentation)


Thanks for the reply.

You say that exploratory studies should be performed to help choose the reaction coordinate. However, how should this be approached? How can I tell if the reaction coordinate is good or bad? Are there any ways of validating the choice of the reaction coordinate?

Best wishes,

hi Dr. alevilla

i was wondering if its possible to pull molecule in negative z direction using gromacs?
in NAMD or amber there is 1 0 -1 functionality but here its just Y/N. please suggest.