[gmx-users] NPT not working

Justin Lemkul jalemkul at vt.edu
Thu Nov 1 21:10:07 CET 2018

On 11/1/18 7:33 AM, Hanin Omar wrote:
> Dear all gromacs user;
> I am new to gromacs, and i desperatly need help. I want to use gromacs to calculate the internal dynamics trajectory within a protein.
> For my simulation i followed this protocol:
> 1)Use pdb2gmx to generate the gro file and topology ( i chose amber99SB and TIP3P water modrel)
> 2)Energy minimization in vacuum
> 3) set perodic boundery coditions
> 4)solvate the system
> 5) ADD ions to neutralize the system
> 6)Energy minimization of the solvated system
> 7) Position restrainted MD
> 8)Unrestraind MD ( NVT equiliberation)
> 9)Unrestraind MD ( NPT equiliberation)
> 10)Run MD simulation with the addition of RDC restraint in the topology like this
> ; Include Position restraint file
> #ifdef POSRES
> #include "RDC-restraint.itp"
> #endif
> The problem is i always get the following eroor when i run the last step
> (Fatal error:
> Error: Too many iterations in routine JACOBI)
> I researched the error and some suggested that the system wasnt equiliberated enough, after checking it seems that the NPT part doesn't converge, i tried extending the time but that didnt solve it, i tired doing it in two steps: 1st NPT with thermostat = v - rescale and barostat = Berendsen for say 6.5 ps and then run 2ND NPT with thermostat = Nose-Hooover and barostat = Parinello -Rahman for 6.5 ps.But that also didnt help. can someone explain why is this happening( and if it has to do with the RDC constraints)? and how can i solve it?

If the introduction of RDC restraints causes a failure, then either (1) 
the implementation of the restraints is incorrect or (2) your structure 
is incompatible with those restraints, leading to the buildup of forces 
and a crash.



Justin A. Lemkul, Ph.D.
Assistant Professor
Virginia Tech Department of Biochemistry

303 Engel Hall
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129


More information about the gromacs.org_gmx-users mailing list