AMBER Archive (2006)

Subject: Re: AMBER: WHATIF Check after MD

From: FyD (fyd_at_u-picardie.fr)
Date: Wed Sep 27 2006 - 23:06:30 CDT

Quoting Fenghui Fan <fenghui_fan_at_yahoo.com>:

> During a MD process, I get several different PDB files
> at different time interal. However when I check these
> MD structures by WHAT IF, I find the structures after
> MD have much poor qualification.
>
> Will you please give me some explainations on these?

I am not surprised by what you report about the quality criterion
reported by whatif, or in fact, whatcheck. My understanding is that
those criteria are specific for globular proteins and not peptides.

Moreover, using whatif, if you build a model by homology modeling for
instance following its own criteria, you will get a model qualified as
'good'.
Did you try to optimize this model ? After few optimization steps the
quality of your model decrease a lot. Consequently, I am not surprised
that MD structures are not in agreement with those criteria.

Now you should check which criteria are bad... Is it the phi/psi
dihedrals or the side chain orientation of some amino acids ? Or bond
length?

I hope this helps, regards, Francois

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