I’m trying to pull a solute (labelled as LIG in the index file) across a lipid bilayer. Below is the pull code part of my mdp file. The problem I’m having at the moment is that every time I try to run the pulling simulation, the solute starts at the centre of the bilayer and not the position I have it in the initial system (positioned on one side of the bilayer). I’ve tried playing around with various different parameters in the mdp, following the approaches I’ve seen in other questions in this forum, but nothing I do changes the fact that the simulation instantly starts by moving the solute to the bilayer centre. I’d really appreciate any help on this because I’m running out of things I can think of trying.
; pull code
pull = yes
pull-coord1-type = umbrella
pull-coord1-geometry = cylinder
pull-coord1-vec = 0.0 0.0 1.0 ; pull along z
pull-cylinder-r = 2.0
pull_nstxout = 100
pull_nstfout = 100
pull-coord1-dim = N N Y
pull_constr_tol = 1e-06
pull_ngroups = 2
pull-group1-name = MEM
pull-group2-name = LIG
pull-coord1-groups = 1 2
pull-coord1-start = no
pull-coord1-rate = 0.0001
pull-coord1-k = 1000
pull-coord1-kB = 1000
This sounds very weird. What do you mean that ‘it starts from the centre’? Does it start there or is it moved there in just a few steps? If the solute is there from step 0 then it was already there when you compiled the binary for the run. Are you by any change compiling with grompp and providing a restraint file with the flag -r and what structure are you providing?
Thanks for getting back to me. The solute starts away from the centre in the input gro file (at the middle x and y coordinates for the box, but about 4 nm away from the surface of the bilayer). It’s there at t=0 in the pulling simulation, but the next frame it’s instantly at the middle z-coordinate for the system (so in the middle of the bilayer). I’m not using a position restraint file when I compile with grompp because I want to solute to move during the simulation, but maybe I should be?
Thank you so much for your help. I have tried pull-coord1-start = yes and had the same problem of the solute moving in the first frame straight to the bilayer centre. I also tried pull-coord1-init set to the intended start position, and various other positions in case I was misunderstanding how the number put for that was interpreted, but so far it hasn’t fixed the issue. But, just to understand your answer better, if I set pull-coord1-start = yes, should it in theory mean that the simulation starts and proceeds from the initial coordinates in the input gro file? If so, could it be an issue with the pull groups (for instance, if I’m pulling the solute, should that be group 1 not 2)? Or, maybe with that knowledge I can go back to a simpler system, starting from scratch, and see if I can get that to work first.
Yes, pull-coord1-start means that it sets the init coordinate to the starting position, so I’m surprised that it’s still getting pulled to the center. Is it possible that you are creating the tpr file with another gro input file than you think?
Since you are pulling with the cylinder geometry it actually matters in which order you specify the pull groups (in pull-coord1-groups). But your settings are correct, the first group is the reference group.
One way to get a better understanding is to turn on the pull-print-com and pull-print-ref-value mdp options. If you then look in the pullx.xvg file you should be able to see the COMs of the two groups, and the target (reference) value, in addition to the current value.
No my point of the restraint file was if you were accidentally using a reference for the restraints with the molecule already in the center of the bilayer, which is the only case where I saw molecules jump very fast to a position. If you are not using any reference, then the -r structure is not the problem. I agree with @MagnusL then that i) it might be that you are already in the center, and ii) printing reference positions will help you. What happens during the trajectory after the jump? The molecule stays in the center or gets dragged out of the bilayer?