Steered MD Simulation using Gromacs

User discussions
, ,
1

GROMACS version:
GROMACS modification: Yes/No
We are attempting to study the transport mechanism of GLUT1 using its natural substrate (glucose) by applying Steered MD simulation. We want to identify the key interactions and residues that are associated with the regulation of the
substrate recognition, binding, translocation, and release.
We set up the system for MD simulation using CHARMM-GUI. A cubic box was chosen as the simulation system. Glucose ligand was found in the vestibule region, according to Direct MD Simulation.
Next, we used the Gromacs system to run the Steered MD Simulation. The glucose molecule is expelled in frame two after the MDP block is executed. We need to re-examine the parameters to ensure that this ligand ejection does not occur.
Enclosing the MDP block we used.
Your suggestions would be of help

; =========================
; Run parameters
; =========================
integrator = md
dt = 0.002
nsteps = 500000
; =========================
; Pulling parameters (SMD)
; =========================
pull = yes
pull_ngroups = 2
pull_ncoords = 1; Pull groups (COM-based)
pull_group1_name = Protein
pull_group2_name = BGLC; Geometry: directional pulling along Z
pull_coord1_type = umbrella
pull_coord1_geometry = direction
pull_coord1_groups = 1 2
pull_coord1_dim = N N Y
pull_coord1_vec = 0 0 1
pull_coord1_start = yes
pull_coord1_k = 500 ; kJ/mol/nm^2
pull_coord1_rate = 0.001 ; nm/ps (1 nm/ns); Reference atom for the Protein pulling group
pull-group1-pbcatom = 3867 ; Replace with the correct atom number
pull-pbc-ref-prev-step-com = yes

; ===== Constraints =====
constraints = h-bonds
constraint_algorithm = lincs

; =========================
; Output control
; =========================
nstxout-compressed = 5000
nstenergy = 1000
nstlog = 1000

2

I see nothing strange in your mdp parameters. I would not use pull-pbc-ref-prev-step-com = yes though. Your force constant is quite low, but that can not cause the expelling.

3

Thank you for your suggestion. We tried by not using pull-pbc-ref-prev-step-com = yes, but still glucose ejects at frame 2. Could there be a possibility that Glucose is present off-axis in the extracellular vestibule of GLUT1. So when we apply directional SMD too early overconstrains ligand motion and causes artificial ejection