[gmx-users] Energy min problem

[Sean Cory]

Hi,

I am running energy minimalization on a protein in water. This protein has 
protonated GLU amino acids (GLUH added with -GLU option) 

When I try to minimize the system, it seems fine at first, but eventually the 
H+ atom begins to move towards the OY atom in the GLU creating huge forces 
dut to the very short distances.

This happens with both CG and STEEP

Does anybody know why this is happening?

Sean


Output of mdrun
*****************************************************************************************************
Polak-Ribiere Conjugate Gradients:
   Tolerance (Fmax)   =  1.00000e+03
   Number of steps    =         2000
   F-max             =  1.98762e+04 on atom 9937
   F-Norm            =  2.31179e+05


Back Off! I just backed up almN6em.trr to ./#almN6em.trr.2#
Step 0, Epot=-1.298480e+05, Fnorm=1.164e+03, Fmax=1.988e+04 (atom 9937)
Step 1, Epot=-1.298484e+05, Fnorm=1.164e+03, Fmax=1.988e+04 (atom 9937)
Step 2, Epot=-1.298490e+05, Fnorm=1.164e+03, Fmax=1.988e+04 (atom 9937)
Step 3, Epot=-1.298500e+05, Fnorm=1.164e+03, Fmax=1.987e+04 (atom 9937)
Step 4, Epot=-1.298516e+05, Fnorm=1.164e+03, Fmax=1.987e+04 (atom 9937)
Step 5, Epot=-1.298541e+05, Fnorm=1.164e+03, Fmax=1.987e+04 (atom 9937)
Step 6, Epot=-1.298583e+05, Fnorm=1.164e+03, Fmax=1.987e+04 (atom 9937)
Step 7, Epot=-1.298650e+05, Fnorm=1.164e+03, Fmax=1.986e+04 (atom 9937)
Step 8, Epot=-1.298758e+05, Fnorm=1.163e+03, Fmax=1.986e+04 (atom 9937)
Step 9, Epot=-1.298933e+05, Fnorm=1.163e+03, Fmax=1.985e+04 (atom 9937)
Step 10, Epot=-1.299217e+05, Fnorm=1.162e+03, Fmax=1.983e+04 (atom 9937)
Step 11, Epot=-1.299674e+05, Fnorm=1.160e+03, Fmax=1.980e+04 (atom 9937)
Step 12, Epot=-1.300411e+05, Fnorm=1.158e+03, Fmax=1.975e+04 (atom 9937)
Step 13, Epot=-1.301583e+05, Fnorm=1.153e+03, Fmax=1.967e+04 (atom 9937)
Step 14, Epot=-1.303483e+05, Fnorm=1.147e+03, Fmax=1.954e+04 (atom 9937)
Step 15, Epot=-1.306512e+05, Fnorm=1.137e+03, Fmax=1.934e+04 (atom 9937)
Step 16, Epot=-1.311296e+05, Fnorm=1.120e+03, Fmax=1.903e+04 (atom 9937)
Step 17, Epot=-1.318727e+05, Fnorm=1.095e+03, Fmax=1.855e+04 (atom 9937)
Step 18, Epot=-1.330096e+05, Fnorm=1.058e+03, Fmax=1.782e+04 (atom 9937)
Step 19, Epot=-1.346865e+05, Fnorm=1.004e+03, Fmax=1.675e+04 (atom 9937)
Step 20, Epot=-1.370650e+05, Fnorm=9.301e+02, Fmax=1.529e+04 (atom 9937)
Step 21, Epot=-1.402525e+05, Fnorm=8.366e+02, Fmax=1.341e+04 (atom 9937)
Step 22, Epot=-1.442533e+05, Fnorm=7.266e+02, Fmax=1.119e+04 (atom 9937)
Step 23, Epot=-1.489093e+05, Fnorm=6.072e+02, Fmax=8.807e+03 (atom 9937)
Step 24, Epot=-1.539434e+05, Fnorm=4.879e+02, Fmax=6.682e+03 (atom 2929)
Step 25, Epot=-1.590008e+05, Fnorm=3.800e+02, Fmax=5.319e+03 (atom 2929)
Step 26, Epot=-1.638971e+05, Fnorm=2.960e+02, Fmax=4.550e+03 (atom 168)
Step 27, Epot=-1.688774e+05, Fnorm=2.407e+02, Fmax=7.004e+03 (atom 168)
Step 28, Epot=-1.747916e+05, Fnorm=3.573e+02, Fmax=4.339e+04 (atom 168)
Step 29, Epot=-1.811522e+05, Fnorm=1.097e+04, Fmax=1.542e+06 (atom 168)
Warning: 1-4 interaction between 151 and 167 at distance larger than 1 nm
These are ignored for the rest of the simulation
turn on -debug for more information

Stepsize too small, or no change in energy.
Converged to machine precision,
but not to the requested precision Fmax < 1000

writing lowest energy coordinates.

Back Off! I just backed up alm_done.gro to ./#alm_done.gro.1#

Polak-Ribiere Conjugate Gradients converged to machine precision in 29 steps,
but did not reach the requested Fmax < 1000.
Potential Energy  = -1.81152183209137e+05
Maximum force     =  1.54217788079911e+06 on atom 168
Norm of force     =  1.09706420361727e+04
***************************************************************************************************

title               =  test1
cpp                 =  /lib/cpp
define              =  -DFLEXIBLE
constraints         =  none
morse               =  no
integrator          =  cg
nsteps              =  2000
nstxtcout           =  1
nstxout             =  1
;
; Energy minimizing stuff
emtol               =  1000
emstep              =  0.000001
;
nstcomm             =  1
ns_type             =  grid
rlist               =  1
rcoulomb            =  1.0
rvdw                =  1.0
Tcoupl              =  no
Pcoupl              =  no
gen_vel             =  no