[gmx-users] problem with dihedral restraints in gromacs 5.0.4

[Justin Lemkul]

On 4/10/15 12:06 PM, Agnieszka Slowicka wrote:
> Hello,
>
> I have a problem with implementation of dihedral restraints in gromacs 5.0.4.
>
> I introduced restraints into a topology file after the section with
> protein topology as fallowing:
>
> ; dihedral restraints added manually: phi,psi, chi1 and chi2
> ; reference values = cluster 1 of charmm/tip4p
> [ dihedral_restraints ]
> ; ai   aj    ak    al     type  phi  dphi  kfac
> ; phi C(n-1) - N - CA - C
>      5    7     9    29     1    -82     0     1
> ; psi N - CA - C - N(n+1)
>      7    9    29    30     1     121    0     1
> ; chi1 N - CA - CB - CG
>      7    9    11    14     1    -177    0     1
> ; chi1 CA - CB - CG - CD1
>      9    11    14     15   1    -106    0     1
>
> The previous version of the topology file with label and power parameters
> doesn't work generating error.
>
> I introduced the restraints section into the .mdp:
>
> ;dihedral restraints
> dihre               =  yes
> dihre_fc            =  100     ; kJ/(mol rad^2)
> dihre_tau           =  0.0
> nstdihreout         =  250
>
>
> Gromacs works with this implementation but the restraints are not take
> into account. I don't see information about input parameters for dihedral
> restraints in my log file, however something is calculated because I found
> there:
>
> Energies (kJ/mol)
>             Bond            U-B    Proper Dih.  Improper Dih.          LJ-14
>      1.71840e+01    6.93870e+01    4.03491e+01    7.20053e+00    4.07828e+01
>       Coulomb-14        LJ (SR)   Coulomb (SR)   Coul. recip.     Dih. Rest.
>      1.80019e+02    6.68397e+03   -4.46824e+04    1.08339e+02    3.82359e-01
>        Potential    Kinetic En.   Total Energy  Conserved En.    Temperature
>
>
> I am sure that my restraints don't work because standard deviation of
> restrained angles is the same order like without restraints.
>

Given that the energy associated with the dihedral restraints is 0.38 kJ/mol, 
I'd say they're just doing very little to help you :)  The dihedral restraint 
options are ignored (well, technically they're removed) in an input .mdp file, 
and it seems that kfac is not a scaling factor at present, it is the actual 
force constant.  That seems to agree with the fact that you're seeing a very 
light restraint with little effect.  Try setting kfac equal to the actual force 
constant you want.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================