Units for the ramd-inputs

[kellegra]

Hi there,

I am familiar to RAMD and NAMD, but wanted to try out the RAMD implementation in GROMACS. I am using GROMACS-2019.4-RAMD-1.0.

In a sample gromacs_ramd.mdp file ramd-r-mind-dist and ramd-max-dist are mentioned to be in nm. When I started the simulation it aborts after a few seconds, telling the displacement of the ligand is greater than the cutoff (distance is 6.8). When I draw the coordinates in PyMOL and measure the distance, these are 6.8 Angstroms. Ramd-max-dist was defined equal to 2.5. So I would guess ramd-max-dist is given in Angstroms not in nm?

Could you please tell me if both, ramd-max-dist and ramd-r-min-dist have to be in Angstroms or nm? And could you also clarify again whether the unit for ramd-force is correct (kJ mol-1 nm-1)?

Best Gerald

[BerndDoser]

Hi Gerald,

thanks for your report. The units in the documentation should be correct (nm and kJ mol-1 nm-1). If I set ramd-max-dist to 4.0 nm the simulation aborts with a receptor-ligand COM distance of 4.02646 nm.

Output:

==== RAMD ==== evaluation 209450
==== RAMD ==== COM ligand position at [6.9189 ,2.13004 ,2.44481]
==== RAMD ==== COM receptor position at [3.68769 ,3.5858 ,4.27185]
==== RAMD ==== Distance between COM of receptor and COM of ligand is 3.98723
==== RAMD ==== Previous COM ligand position at 6.88821 ,2.15077 ,2.452]
==== RAMD ==== Previous COM receptor position at 3.68967 ,3.58652 ,4.2719]
==== RAMD ==== Change in receptor-ligand distance since last RAMD evaluation is 0.0389704
==== RAMD ==== evaluation 209500
==== RAMD ==== COM ligand position at [6.95324 ,2.11111 ,2.43888]
==== RAMD ==== COM receptor position at [3.68658 ,3.58744 ,4.27239]
==== RAMD ==== Distance between COM of receptor and COM of ligand is 4.02646
==== RAMD ==== Maximal distance between ligand and receptor COM is reached.
==== RAMD ==== GROMACS will be stopped after 209500 steps.

Best regards, Bernd

[kellegra]

Hey Bernd, thx for your reply! My problem that the RAMD simulation stops after 50 steps of simulation time still remains. I followed the tutorial on kbbox. I decided to start from a new NVT simulation due to something could have went wrong prevously.

NVT and NPT runs look fine so far. When I start the RAMD run it aborts after 50 steps, cause the distance is > 6 nm. Threshold is 2.5 nm. Then I had a look at the last RAMD frame and the COMs were 1.1 nm apart only?!

What am I overlooking here? If it makes things easier we could change to German.

Output:

``==== RAMD ==== evaluation 50
==== RAMD ==== COM ligand position at [2.42658 ,1.2378 ,5.70149]
==== RAMD ==== COM receptor position at [2.01008 ,7.77483 ,4.9308]
==== RAMD ==== Distance between COM of receptor and COM of ligand is 6.59546
==== RAMD ==== Maximal distance between ligand and receptor COM is reached.
==== RAMD ==== GROMACS will be stopped after 50 steps.

.mdp file:

`integrator                       = md
dt                              = 0.002
nsteps                           = 20000000

nstlog                          = 2500
nstenergy                        = 2500
nstxout-compressed               = 500

continuation                     = yes
constraints                      = h-bonds
constraint-algorithm             = lincs

cutoff-scheme                   = Verlet
coulombtype                     = PME
rcoulomb                        = 1.0

vdwtype                         = Cut-off
DispCorr                        = EnerPres

ramd                            = yes
ramd-seed                       = 981   ;  XX will be  replaced by the trajectory number
ramd-force                      = 585.2      ; 585.2 kJ mol-1 nm-1  =  14 kcal/mol A *10*4.18
ramd-eval-freq                  = 50  ; number of steps in each RAMD cycle
ramd-r-min-dist                 = 0.0025       ; nm
ramd-force-out-freq              = 10
ramd-max-dist                   = 2.5         ; nm
ramd-receptor                   = Protein
ramd-ligand                      = LIG

comm-mode                        = Linear
nstcomm                         = 100
comm_grps                       = System

tinit                           = 0.000
nstxout                         = 5000
nstvout                          = 5000
compressed-x-precision   = 500
compressed-x-grps                = SYSTEM

pbc                             = xyz
rlist                           = 1.0
fourierspacing                  = 0.12
pme_order                       = 4
ewald_geometry                  = 3d
ewald-rtol                       = 1e-5
ewald-rtol-lj                    = 1e-5
vdw-modifier                     = Potential-shift
rvdw-switch                      = 0.0
rvdw                             = 1.0

tcoupl                          = Nose-Hoover
tc_grps                         = Protein_LIG_NAP    Water_and_ions
tau_t                    = 1.0    1.0
ref_t                    = 300 300

Pcoupl                           = Parrinello-Rahman
pcoupltype                       = isotropic
tau_p                    = 5.0
compressibility                 = 4.5e-5
ref_p                           = 1.0
gen_vel                          = no

` Best Gerald

[BerndDoser]

Hi Gerald,

laut Ausgabe sind die COM Positionen [2.42658, 1.2378, 5.70149] fuer den Liganden und [2.01008, 7.77483, 4.9308], was einem Abstand von 6.6 nm entspricht. Wie bestimmst Du die COM Positionen? Du kannst mir gerne Deine Input-Dateien schicken. Dann kann ich mal drueberschauen.

Gruss Bernd

[BerndDoser]

Source of this issue is #18.