Choose proper awh1-dim1-force-constant and -diffusion in AWH

User discussions
1

GROMACS version: 2024.5
GROMACS modification: No
Hi!

I am studying the conformational transition of a two-component signaling system upon ligand binding, specifically how the structural changes enable histidine kinase to undergo correct phosphorylation.

Here is my mdp:
integrator = md
dt = 0.002
nsteps = 100000000
nstxout-compressed = 25000
nstxout = 0
nstvout = 0
nstfout = 0
nstcalcenergy = 200
nstenergy = 10000
nstlog = 10000
;
cutoff-scheme = Verlet
nstlist = 20
rlist = 0.9
vdwtype = Cut-off
vdw-modifier = None
DispCorr = EnerPres
rvdw = 0.9
coulombtype = PME
rcoulomb = 0.9
;
tcoupl = v-rescale
tc_grps = SOLU MEMB SOLV
tau_t = 1.0 1.0 1.0
ref_t = 303.15 303.15 303.15
;
pcoupl = C-rescale
pcoupltype = semiisotropic
tau_p = 5.0
compressibility = 4.5e-5 4.5e-5
ref_p = 1.0 1.0
;
constraints = h-bonds
constraint_algorithm = LINCS
continuation = yes
;
nstcomm = 100
comm_mode = linear
comm_grps = SOLU_MEMB SOLV
;
pull = yes
pull-ncoords = 2
pull-ngroups = 4
pull-nstxout = 1000
pull-group1-name = group1
pull-group2-name = group2
pull-group3-name = group4
pull-group4-name = group6
;
pull-coord1-groups = 1 2
pull-coord1-geometry = distance
pull-coord1-type = external-potential
pull-coord1-potential-provider = AWH
;
pull-coord2-groups = 3 4
pull-coord2-geometry = distance
pull-coord2-type = external-potential
pull-coord2-potential-provider = AWH
;
awh = yes
awh-nstout = 20000
awh-nbias = 1
awh1-ndim = 2
;
awh1-dim1-coord-index = 1
awh1-dim1-start = 0.3
awh1-dim1-end = 5.3
awh1-dim1-force-constant = 100000
awh1-dim1-diffusion = 5e-3
;
awh1-dim2-coord-index = 2
awh1-dim2-start = 0.3
awh1-dim2-end = 5.3
awh1-dim2-force-constant = 100000
awh1-dim2-diffusion = 5e-3

In this process, I am using AWH and I am unsure whether the awh1-dim1-force constant and -diffusion values I chose are appropriate. Should these parameters be specifically calculated based on my system, or is it reasonable to use a more general value?

Any insights or recommendations would be greatly appreciated. Thanks!

2

The diffusion is rather uncritical and only matter in combination with the initial error, read the manual.

The force constant is important to get right. This should be about an order of magnitude larger than the curvature of the free-energy landscape. Choosing is too large can lead to integration issues. If you have no idea what it should be, you can experiment with some short runs.

3

Hi hess,

Thank you very much for your reply (sorry for being ghost for 3 month as I was working in labs)! I have two results here, with 80000 and 60000 force constant. Actually I am not quite sure what data it shows. Does the low energy area means a low energy conformation at this coordinate? And how to distinguish proper force constant?

image
image1483×598 122 KB

4

A force constant of 80000 or 60000 is identical in practice. You free-energy values (I don’t know the unit) are extremely high. So your results are probably far from converged and rather meaningless.

5

Thank you for your reply. The unit is kJ/mol. How should I determine the curvature of the free-energy landscape? And should I run it longer to make it converged? Thank you again!

6

At the moment there is not much to say as you are far from convergence. You might need a factor 10, 100 or 1000 longer sampling.

To determine what force constant you need, you need to check whether the coordinate and reference distribution match sufficiently well. But you might need more sampling for that.

7

Thank you Hess. How can I check the mean number of samples?