[gmx-users] Applying Dihedral Restraints to Polymer in GROMACS 5.1

[Harry Ponce]

Hello,

I'm attempting to run an energy minimization on some high energy conformers
of a polymer dimer while restraining two dihedral angles pertaining to the
five backbone carbons of interest. The literature and forums online are
sparse when it comes to non-biochemical systems, and it seems as though the
syntax might have changed over the updated versions of GROMACS.

I've attempted to implement this bit in my topology:

#ifdef DIHRE

[ dihedral_restraints ]
; ai   aj    ak    al  type  label  phi  dphi  kfac  power
    2    1     8    29     1      1  52      0     1      2
    3    2     1     8     1      1  75      0     1      2

#endif

As well as this bit in my .mdp file:

define            = -DDIHRE
; Dihedral Restraints
dihre            = yes
dihre_fc        = 1000
dihre_tau        = 0.0
nstdihreout        = 50

I've also tried using what I think is the more recent syntax for applying
restraints:

; position restraints for System of UNNAMED

[ position_restraints ]
;  i funct       fcx        fcy        fcz
   1    1       1000       1000       1000
   2    1       1000       1000       1000
   3    1       1000       1000       1000
   8    1       1000       1000       1000
  29    1       1000       1000       1000

However, when I energy minimize and read the backbone dihedral angles they
are different than the ones I started with! I am sure I have the correct
atom numbers (1,2,3,8,29), but am not sure why GROMACS is reading them and
then ignoring my restraints.

Any ideas?

Thank you very much,
Harrison