Pair list generation problem in GROMOS96
- From: Mateusz Nowak <mynowak-: at :-cyf-kr.edu.pl>
- Subject: Pair list generation problem in GROMOS96
- Date: Tue, 13 Feb 2001 11:09:10 +0100 (MET)
Dear CCL
I'm trying to run a regular MD simulation in GROMOS96. My system is 107 aa
protein in a box of SPC water with 4 NA+ ions added. PBC are applied, and
the box sizes are ~ 4*5*6 nm. Bonds involving hydrogens are shaken. In an
output i'm getting following warnings:
NONBML: WARNING, NOT ALL EXCLUDED NEIGHBOUR WERE SCANNED IN THE PAIR-LIST
NONBML: WARNING, NOT ALL THIRD NEIGHBOUR WERE SCANNED IN THE PAIR-LIST
Kinetic and non-bonded potental energy terms are not being calculated
("nan" instead of a value) in an initial (t=0) step. In next steps
pair
list contains 0 elements instead of ~300000, as it is in initial step.
According to manual this warning occures when cut-off for pair list
construction (RCUTP) is not reasonably large. In my case it has 1.0 nm
(acctually twin rang cut-off 1.0/1.6 is used). I tried to run the same
simulation with pair list construction cut-off 1.2 and 1.6 nm, but the
same warning occures.
Has anybody experienced similar problem? Any help will be appreciated.
Mateusz Nowak
Jagiellonian Univ.
Krakow
POLAND