AMBER Archive (2002)

Subject: Re: Carnal coordinate averaging

From: Bill Ross (photoriot_at_yahoo.com)
Date: Mon Feb 25 2002 - 16:44:17 CST

> In Amber 5, when I try to create a
> coordinate-averaged PDB file for all or
> part of an MD simulation, I get a structure
> with bond lengths that are obviously wrong,
 
How obviously? E.g. methyl H's will average to a
short bond. The normal thing is to energy minimize
the average. Of course waters need to be ignored
altogether as the average bonds can interpenetrate
the structures, so they need to be stripped. It's
also a good idea to RMS FIT the trajectory to keep
things as aligned as possible.

> I realize that carnal is averaging the waters as
> well, and tried to declare a group consisting of
only
> the solute to avoid this, but the program seemed not

> to recognize the declaration.

Example input/output would be useful here.

> Is there a more recent version of carnal,
> analogous to the new ptraj (6.3), that would work
> better for averaging?

The latest 6.0 versions of carnal & leap can be
downloaded from the amber web.

Bill Ross

FYI - my regular email is ross_at_cgl.ucsf.edu - I don't
read this yahoo address as often.

__________________________________________________
Do You Yahoo!?
Yahoo! Sports - Coverage of the 2002 Olympic Games
http://sports.yahoo.com