[gmx-users] Test particle insertion in designated coordinates gives Segmentation fault

[Zheng, Lingru]

Hello everyone,

I have a problem using tpic. Basically following the process here: https://groups.google.com/forum/#!topic/archive-gmx-users/hh2cq9ZtqgU

My system is a surface in contact with liquid water and within which there are some ions to balance the surface charge. There is a few nm space above the liquid water layer. I intend to insert water molecule into the water liquid-vapour interface. Here are my steps:

1. I equilibrated the whole system using NVT for 40ns, then use trjconv -sep to split the .trr file into a series of configuration files conf*.gro

2. I use a script to add the following cavity coordinates to the bottom of each of the conf*.gro file:
    601INS      X 3551   2.565   2.196   5.050

3. I then load all of the conf*.gro into vmd, and 'save coordinates' to obtain a new trajectory: nvt_cavity.trr

4. I then modify the final configuration (.gro) from the NVT run by adding the following lines at the bottom, and save it as nvt_cavity.gro :
  601SOL      OW 3551   0.028  -0.033   0.000
  601SOL    HW1 3552  -0.072  -0.033   0.000
  601SOL    HW2 3553   0.044   0.066   0.000
  601INS           X 3554   2.565   2.196   5.050

5. I added an insert.itp file to the forcefield folder as otherwise gromacs cannot recognise the atom/molecule, it contains:
[ moleculetype ]
; molname nrexcl
INS         1

[ atoms ]
; id    at type res nr  residu name at name  cg nr  charge
1        X      1      INS           X       1      0

6. I also add the following to the ffnonbonded.itp file (Here I am unsure)
[ atomtypes ]
         X   9      18.00000     0.0000 A       0.000000    0.000000

7. To the .top file I added two lines to the [molecules]:
SOL         1
INS          1

8. Then I run the following:
gmx grompp -f tpi.mdp -o tpi -pp tpi -c nvt_cavity.gro -p topol.top<http://topol.top>
gmx mdrun -v -deffnm tpi -rerun nvt_cavity.trr -tpid energy

It generates the .tpr file but gives 'Segmentation fault’. I do not understand why. Can anyone please help?

Below is my tpi.mdp file:

; Run parameters
integrator = tpic ; leap-frog integrator
nsteps = 10000 ; run these many tpi's
dt              = 0.01
nstlist = 10 ; number of attempted insertions each nstep
rtpi = 0.05 ; radius of sphere for insertions (nm)

; Neighborsearching
cutoff-scheme   = group ; verlet scheme not valid for tpi
pbc = xyz ; 3-D PBC
rlist = 1.4 ; short-range neighborlist cutoff (in nm)

; Electrostatics and vdw
coulombtype = PME ; Particle Mesh Ewald for long-range electrostatics
vdwtype = cutoff
rcoulomb = 1.4 ; short-range electrostatic cutoff (in nm)
rvdw = 1.4 ; short-range van der Waals cutoff (in nm)
DispCorr = EnerPres        ; account for cut-off vdW scheme


I previously ran the tpi algorithm, it works  but the excess chemical potential varies significantly with nlist and nstep even for a 40ns length trajectory, so I am trying to use tpic to see if it would give more reliable results.

Thanks in advance for your help!
Lynn