Dear all

I have few confusions about the pulling code in gromacs, I was using NAMD before and the method was quite straighforward in NAMD to pull the ligand out of the protein, in gromacs it is also straight forward but I need to clear few things

1. In NAMD one can define the pulling direction by using SMDDir in the input file. Difference between the COM distance of SMD atom and fixed atom gives the pulling direction, SMD atoms are usually the heavy atoms of the ligand whereas the CA backbone of the protein is kept fixed. This difference between the COM distances of both tells the pulling direction of the ligand in NAMD in case one does not know the pulling direction of the ligand. Is pull-coord1-vec is same as NAMD SMDDir ? And in NAMD you define the SMD atoms and fixed atoms, here in Gromacs I am confused how it defines the fixed atoms or reference atom which is protein from where ligand has to be pulled out?

I followed the umbrella sampling tutorial in Gromacs

2. Secondly I did few tests regrading pulling out the ligand from the protein using the pull code as below, this is just a test code to understand things thats why applied quite fast velocity.

; Pull code

pull = yes

pull_ncoords = 1 ; only one reaction coordinate

pull_ngroups = 2 ; two groups defining one reaction coordinate

pull_group1_name = Ligand

pull_group2_name = Protein

pull_coord1_type = umbrella ; harmonic potential

pull_coord1_geometry = distance ;

pull-coord1-vec = -0.49 -0.85 0.16 ;

; pull-coord1-origin = 53.12 52.96 53.066

pull_coord1_dim = Y Y Y ;

pull_coord1_groups = 1 2 ;

pull-pbc-ref-prev-step-com = YES

pull-group1-pbcatom = 0

;pull-group2-pbcatom = 0

pull_coord1_start = yes ; define initial COM distance > 0

pull_coord1_rate = 0.01 ; 0.0004 nm per ps

pull_coord1_k = 600 ; kJ mol^-1 nm^-2

I changed one parameter of the pulling code every time to see whether there is any affect on the pulling but by visualizing the trajectories I didn't see any difference, the ligand is still coming out even if I switch off pull-coord1-vec, pull_coord1_dim together. Also with these feature if I changed distance to direction still the ligand comes out.

Secondly if I just keep pull_coord1_dim Y Y Y in all three dimensions it still takes the ligand out of the protein in almost similair direction as given by pull-coord1-vec, if I dont give pull-coord1 vec and only give pull coord1 dim Y Y Y will my ligand be pulled in a random direction?

Also What is pull-coord1-origin? is it the origin of the box?

Anyone kindly let me know what changes should I make in the above given code where I just want my ligand to come out of the protein in a random pulling direction or by the method of NAMD SMDDir?

I am mostly interested in the pulling forces required to pull the ligand, so what should be the pulling code input in gromacs for that?

Thank you in advance, your answer will help me in understanding the pulling code and applying it to my work!!