Umbrella sampling of Lipid Protein system using Martini FF

User discussions
1

GROMACS version: 2020.4
GROMACS modification: No

Hello,
I am calculating potential of mean force (PMF) for a protein-lipid bilayer system using Coarse Grained Martini force field and Gromacs 2020.4. In steered MD simulation I applied force to get a reaction coordinates in z-direction, where protein moved closer to the lipid bilayer and reached to the bilayer center (using pull code of Gromacs).
I generated a series of initial configurations, each corresponding to a location where the protein was harmonically restrained at decreasing center-of­-mass (COM) distance from the lipid molecule (from 3.9 nm to 0.5 nm) using an umbrella biasing potential following Dr. Justin A. Lemkul tutorial.
For Umbrella sampling (US), I generated a total of 37 different starting configurations along the z-reaction coordinate. Each configurations were independently equilibrated followed by the US for 10ns using following pull code:

; Pull code
pull = yes
pull_ncoords = 1 ; only one reaction coordinate
pull_ngroups = 2 ; two groups defining one reaction coordinate
pull_group1_name = PROTEIN
pull_group2_name = MEMBRANE
pull_coord1_type = umbrella ; harmonic potential
pull_coord1_geometry = distance ; simple distance increase
pull_coord1_dim = N N Y ; pull along z
pull_coord1_groups = 1 2 ; groups 1 and 2 define the reaction coordinate
pull_coord1_start = yes ; define initial COM distance > 0
pull_coord1_rate = 0.0 ; 0.001 nm per ps = 1 nm per ns
pull_coord1_k = 1000 ; kJ mol^-1 nm^-2
pull_coord1_init = 0.2

pull_group1_pbcatom = 20
pull-pbc-ref-prev-step-com = yes

This is generating 37 trajectories, each of 10 ns. To use in WHAM I am using the last 5ns from each trajectory. Everything seems to be running fine, however, I am getting histogram, which seems not to be correct (see attachment please), as each window is not overlapping.
I have three questions:
Q.1 Why my COM distance is increased form 0.5 nm to over 1nm during production run despite using pulling force?
Q.2 Why windows in my histogram is not overlapping? If I use small force constant, the COM starts increasing.
Q.3 How can I resolve it?

Many thanks for taking time and reading the message. Highly appreciated. Looking forward for the solution.

histo_dist_US
histo_dist_US792×612 74.1 KB

2

Hi,

The lack of overlapping between each window means that you did not reach convergence. A option is to expend the run and check for convergence.

Concerning your second issue, according to mdp file you define the distance as COM of starting conformation at t=0 + 0.2 nm)

pull_coord1_start = yes => add the COM distance of the starting conformation to pull-coord1-init
pull_coord1_init = 0.2

I am not sure that it is what you want. Using this setting, to have to sure what is the distance at t=0. One possible alternative is to use pull_coord1_start = no and define pull_coord1_init = selected distance.

Kind regards
Alessandra