[gmx-users] AMBCONV tip3p problem

[Berk Hess]

Hi,

This is a message for people who use the AMBCONV package from
the Gromacs contributions page.
The tip3p.itp file included in this package is not correct.
It uses settle constraint as well as normal bonds and an angle.
When used in a run with constraints=all-bonds, the number of degrees of
freedom for one tip3p molecule is counted as 4 instead of 6.
This means the temperature is 50% higher than reported.
Thus runs with this tip3p.itp and constraint=all-bonds are incorrect.

To correct this the [ bonds ] and [ angles ] sections in tip3p.itp should be 
removed.

Berk.

_________________________________________________________________
Play online games with your friends with MSN Messenger 
http://messenger.msn.nl/