AMBER Archive (2004)

Subject: AMBER: unformatted md problem

From: richard dimelow (r.dimelow_at_stud.man.ac.uk)
Date: Mon Sep 27 2004 - 08:13:40 CDT

I'm having a problem with a molecular dynamics run in sander. The input file
i'm
using is

  ntx=4, irest=1, ntxo=0,
  imin=0,
  ntpr=10,
  cut=999,
  ntf=2, ntc=2,
  ntb=0,
  nstlim=1000,
  igb=0

The coordinates and velocities are read unformatted (ntx=4), i am not using
periodic boundary conditions (ntb=0) and i am not using the generalised born
model (igb=0). When doing this the following error message appears in the
output

| Flags: SGIFFT MEM_ALLOC HAS_FTN_ERFC
| NONPERIODIC ntb=0 and igb=0: Setting up nonperiodic simulation
 Cannot read unformatted coord for nonperiodic system
 For ntb=0, ntx must be 1,5,7,9

The code which prints out this message is in ew_box.f (line 850), which
is confusing since i am not using periodic boundary conditions and so there
shouldn't be a need for box information. What's more, if i read the
coordinates
and velocities formatted using ntx=5 then there is no problem.
Alternatively, if I
keep ntx=4 but use the generalsed born model igb=1 then again there is no
problem.

So does anyone know if there a physical reason why I can't use ntx=4 with
igb=0, or have I made a mistake with my input file. Cheers.

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