[gmx-users] Fwd: restraints failure problem

Sadaf Rani sadafrani6 at gmail.com
Tue Feb 25 01:04:08 CET 2020


Dear Gromacs users

I am posting this problem third time in hope to get some suggestions
regarding my problem.
I am doing a test calculation in MD simulation between ligand and a protein
residue for bond angle and dihedral restraints. I have set them in my
topology file as below:-
; distance restraints
[ bonds ]
;    i        j     type     r0A     r1A     r2A    fcA    r0B     r1B
r2B    fcB
    71    29    10     0.403   0.403   10.0   0.0    0.403   0.403   10.0
4184.000

[ angle_restraints ]
;   ai    aj    ak    al  type    thA      fcA    multA  thB      fcB
 multB
    29    71    69    71   1  104.92   0.0    1     104.92   41.840   1
    71    29    31    29   1  147.02   0.0    1     147.02   41.840   1

[ dihedral_restraints ]
;   ai    aj    ak    al   type    phiA     dphiA  fcA    phiB      dphiB
 fcB
    69    71    29    31   1    146.72     0.0    0.0   146.72     0.0
41.840
    72    71 29    31   1     17.55      0.0    0.0    17.55     0.0
41.840
    72    71 29    25   1     173.10     0.0    0.0   173.10     0.0
41.840

However, after energy minimization, I observed that my restraints are not
working and I find the following change:-

bonds changed from 0.403 *0.401*
angles:-
104.92   changed to *107.98*
147.02  changed to *152.76*
dihedrals:-
 146.72  changed to *141.72*
 17.55  changed to 2*2.13*
173.10  changed to *174.23*
  The fluctuation in restraints goes on increasing till the production run
which shows that my restraints are not working.
How should I work with these so that they may not change much? Do I need to
increase force for restraints if yes then how much increase? any idea?  I
have tried these simulations with position restraints. Is the method of
putting restraints correct or should I try some other? I have to apply
these in the free energy calculation which already crashed because of
restraint failure.
I would really be thankful for your kind suggestions, please.

Thanks.
Sadaf


More information about the gromacs.org_gmx-users mailing list