AMBER Archive (2004)

Subject: AMBER: segmentation fault with sander

From: Carsten Detering (detering_at_u.washington.edu)
Date: Tue Mar 16 2004 - 20:29:55 CST

Hi Ambers,

I am getting a segmentation fault after sander has been computing for
approx. 1 minute, it doesnt matter whether I use both processors (I had
posted a message about a week ago that a 'Child process died
unexpectedly from signal 11' and mpirun crashed')

I am calculating a protein+ligand complex, this time without solvent. I
created a box in xleap (setbox unit vdw), saved amberparm. The idea was
to first get a good starting structure, before I solvate it. I use this
input file:

&cntrl
imin=1, maxcyc = 10000
ntpr = 5
cut=10.0
ntb = 1
&end

This is the last step of the calculation:
NSTEP ENERGY RMS GMAX NAME NUMBER
     380 -4.1232E+06 2.2389E+09 1.6665E+11 HO'2 4661

  BOND = 215.1359 ANGLE = 817.1759 DIHED =
2613.2389
  VDWAALS = -2570.2908 EEL = -21138.6886 HBOND =
0.0000
  1-4 VDW = 1011.9867 1-4 EEL = -4104146.4708 RESTRAINT =
0.0000

I realize that I have quite a high 1-4 electrostatics value, but I don't
know how to get rid of that. I set all my histidines protonated (HIP),
and my overall charge on my protein is +6.
I also tried to get rid of close contacts in xleap by minimizing the
respective residues (with Edit/relax selection), but that didn;t help.
They are still listed when I check my unit.

Does anyone have an idea why my system always crashes? Before, it
finished the calculation (with solvent, though) and I got at least some
bad energy values...

Thanks in advance,

Carsten 

-- 

Carsten Detering
University of Washington
Seattle, WA 98195
Fon 206.543.5081
Fax 206.685.8665

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu