[gmx-users] Normal Mode Analysis with Gromacs 4.6

[mcfc1301]

Hello Everyone，

I'm confused about Normal Mode Analysis of Gromacs. I have googled a lot but
find little useful.
My Gromacs was installed in a Linux cluster.

Follows is my steps to do normal mode analysis:

• pdb2gmx -ignh -ff < forcefield > -f 1SHS mutated.pdb -o mut.gro -p
topol.top
• editconf -f mut.gro -o box.gro -d < spacingdistance >
• grompp -f em.mdp -c box.gro -p topol.top -o em.tpr
• mdrun -s em.tpr -deffnm em -c em.g96 -v
• grompp -f nm.mdp -c em.g96 -o nm.tpr 
  /--------------->WARNING 1 [file nm.mdp]:
                        The switching range for PME-Switch should be 5% or
less, energy
                        conservation will be good anyhow, since ewald_rtol =
1e-05/

I have taken advises from the forum that I should not genbox and should use
cg instead steep for em in order to do Normal Mode Analysis, but now I am
facing problem of PME-Switch.

Does anyone have any experience on this?
         
Thanks in advance.

Kevin


==============Attached is em.dmp and nm.dmp=========================
*em.mdp
*
define              =  -DFLEXIBLE
constraints         =  none
integrator          =  cg
dt                  =  0.002    ; ps !
nsteps              =  400
nstlist             =  10
ns_type             =  grid
rlist               =  1.0
coulombtype         =  PME
rcoulomb            =  1.0
vdwtype             =  cut-off
rvdw                =  1.4
optimize_fft        =  yes
emtol               =  1000
emstep              =  0.01

---------------------------------------
*nm.mdp*
constraints         =  none
integrator          =  nm
dt                  =  0.002    ; ps !
nsteps              =  400
nstlist             =  10
ns_type             =  grid
rlist               =  1.2
coulombtype         =  PME-Switch
rcoulomb            =  1.0
vdwtype             =  Cut-off
rvdw                =  1.4

--
View this message in context: http://gromacs.5086.x6.nabble.com/Normal-Mode-Analysis-with-Gromacs-4-6-tp5013240.html
Sent from the GROMACS Users Forum mailing list archive at Nabble.com.