[gmx-users] GROMACS 5 - problem with LINCS with freeze groups

[Mark Abraham]

Hi,

You are using settings that trigger the group-scheme twin-range code.
There's been a code bug there for a very long time, which we fixed after
4.6.5 (see http://redmine.gromacs.org/issues/1400. "Normal" usage such as
in your .mdp file also tended to be broken when simulating water, which
masked the code bug; your 4.6.5 run may look smooth, but probably the
thermostat is hiding the systematic problem of nstlist*dt being too large
to properly model water libration, and the symptoms of the virial problem
we fixed are lost in the noise.

If you can try 4.6.7, I expect that you will observe the same problems you
see in 5.0.x. That will help us know what to consider fixing. We certainly
didn't test the freeze-group functionality, so we could well have
introduced a new problem, such as adding force components that were already
zeroed for freezing. If I'm right about 4.6.7 also being broken, please
file an issue at http://redmine.gromacs.org with a tarball of your grompp
input files.

Mark

On Tue, Nov 11, 2014 at 2:05 PM, Tomek Wlodarski <tomek.wlodarski at gmail.com>
wrote:

> Hi,
>
> I experience some strange behaviour and I wonder if this is a bug or I am
> doing something wrong.
> I am running test simulation to check freeze groups in new gromacs.
> And I am getting LINCS Warnings for atoms which should be frozen...:
>
> Step 0, time 0 (ps)  LINCS WARNING
> relative constraint deviation after LINCS:
> rms nan, max inf (between atoms 30 and 31)
> bonds that rotated more than 30 degrees:
>  atom 1 atom 2  angle  previous, current, constraint length
>      32     33   90.8    0.1010  23.4699      0.1010
>      32     34   90.3    0.1010  71.2488      0.1010
> [...]
>
> my mdp file:
>
> ; RUN CONTROL PARAMETERS
> integrator               = md
> dt                       = 0.002   ; 2 femtosecond time step for
> integration
> nsteps                   = 500000  ; 1ns
>
> ; OUTPUT CONTROL OPTIONS
> nstxout              = 5000    ; save coordinates every 2 ps
> nstvout              = 5000    ; save velocities every 2 ps
> nstenergy             = 5000    ; save energies every 2 ps
> nstlog           = 5000    ; update log file every 2 ps
> energygrps           = Protein Non-Protein
> nstxtcout                = 5000
> xtc-grps                 = non-Water
>
> ; NEIGHBORSEARCHING PARAMETERS
> nstlist                  = 5
> ns-type                  = Grid
> cutoff-scheme            = group
> pbc                      = xyz
> rlist                    = 0.9
>
> ; OPTIONS FOR ELECTROSTATICS AND VDW
> coulombtype              = pme
> rcoulomb                 = 0.9
> fourierspacing       = 0.12
> ewald_rtol         = 1e-5
> fourier_nx               = 0
> fourier_ny               = 0
> fourier_nz               = 0
> optimize_fft         = yes
> vdw-type                 = Cut-off
> rvdw                     = 1.2
> energygrp_table         =
>
> ; Temperature coupling
> Tcoupl                   = V-rescale
> tc-grps                  = Protein    Non-Protein
> tau_t                    = 0.1        0.1
> ref_t                    = 300        300
>
> ; Freeze atoms
> freezegrps               = Protein
> freezedim                = Y Y Y
> energygrp-excl           = Protein Protein
>
> ; Pressure coupling
> pcoupl                   = no
>
> ;Velocity generation
> gen_vel          = no
> gen_temp         = 300
> gen_seed         = -1
>
> ;Constrain all bonds
> constraints         = all-bonds
> constraint-algorithm     = Lincs
> lincs_iter              = 2             ; accuracy of LINCS
> lincs_order          = 6            ; also related to accuracy
>
> Exactly the same system with the same mdp file I am running in Gromacs
> 4.6.5 and everything runs smoothly.
> Any suggestions?
> Thanks a lot!
> Best
>
> tomek
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>