GROMACS version: 2022.2
GROMACS modification: Yes/No
Here post your question
I have tried to pull one atom out of a big structure. The connection between the atom and the big structure is not a covalent bond, just a coordinate bond. I use the following settings for pulling:
Pull code
pull = yes
pull_ncoords = 1
pull_ngroups = 2
pull_group1_name = Al20
pull_group2_name = MOF_Al20
pull_coord1_type = umbrella
pull_coord1_geometry = distance
pull_coord1_groups = 1 2
pull_coord1_rate = 0.05
pull_coord1_k = 1500
pull-group2-pbcatom = 6385
pull-pbc-ref-prev-step-com = yes
Afterward, I use “gmx energy” command to extract the energy information (the options of energy has shown in the following). I want to know restraining energy, the one that is used to bring the atom to the desired position. I am wondering if the “COM-pull energy” is the one I needed. I did not find any definition of COM-pull energy in gromacs manual.
COM-Pull energy is the energy of the harmonic potential used for pulling the group.
But this energy is not very meaningful. Usually you want to have the potential of mean force, which is the integral of the mean pulling force over distance.
Thank you for your feedback. The reason that I am checking the pulling process is the PMF after umbrella sampling does not make sense.
The PMF and histograms and the non-bonded interactions between my pulling atom and the big structure have shown below. In principle, my structure should be stable at the beginning and then will be disturbed by the pulling force. However, the PMF I obtained by “gmx wham it tpr-files.dat -if pullf-files.dat -o -hist -unit Cal” does not show as expected, and a minimum energy state is obtained when the distance between the atom and the structure is far away 1.7nm. I have added 20 windows for umbrella sampling between the distance 0 - .5nm, the PMF showed similar trend, and the histograms for this range are not attached here. By checking the COM-Pull energy, I can see it showed a quite different pattern.
Could you figure out what is problem with my umbrella sampling for the energy barrier calculation? Why the minimum value is around 1.7nm, what is wrong with my PMF analysis?
Yes, the minimum value should at the beginning, not be far away.
In my figure, the PMF shows a decreasing trend and reaches a minimum value of around 1.7 nm, which seems not correct. What I expect is a minimum value is around 0.2 nm (the common distance of the coordinate bond between the pulling atom and the structure) and then increases until reach a plateau far away. The decreasing trend before 1.7 nm and the minimum value at 1.7 nm seems not right.
Yes, what I expect is a minimum value of PMF value at the very beginning (the original distance between my pulling atom and the coordinate atom in the big crystalline structure is 0.2 nm) and then start to increase to reach a plateau. My trouble is I have no idea where I made mistakes and get this weird result. It seems that my pulling process is okay. The production simulation is NVT equilibrium (100ps) followed by 1ns production simulations. My pull-related parameters are as below:
Pull code
pull = yes
pull_ncoords = 1
pull_ngroups = 2
pull_group1_name = Al20
pull_group2_name = MOF_Al20
pull_coord1_type = umbrella
pull_coord1_geometry = distance
pull_coord1_groups = 1 2
pull-coord1-init = 0.2 (the actual value of the distance obtained from the pulling process)
pull_coord1_rate = 0.00
pull_coord1_k = 1500
pull-group2-pbcatom = 6385
pull-pbc-ref-prev-step-com = yes
Then I use the command " gmx wham it tpr-files.dat -if pullf-files.dat -o -hist -unit kCal" to get the PMF and histograms.
Can you figure out what is wrong? What can I try to solve this problem?
One thing that might be an issue is the geometry. With geometry distance the atom can go in any direction from the COM of the structure. I would usually set up a periodic substrate in the X/Y plane and use geometry direction along Z. I don’t know if you can use such a setup.
