AMBER Archive (2005)

Subject: Re: AMBER: inconsistent ordering in an improper dihedral of TRP

From: Bill Ross (ross_at_cgl.ucsf.edu)
Date: Fri Feb 11 2005 - 19:56:10 CST


>> When there are more than two TRP residues, the first one follows the
>> order CD1-CD2-CG-CB and all the others CD2-CB-CG-CD1
>
> I can confirm that this happens when I try multiple TRP too. I don't
> know why- I'll have to look at leap but maybe someone more familiar with
> the leap torsion builder would be helpful too..

To my mind, it's the hardest code in leap. It might be worth
testing if this happens to other residues as well.

I believe this difference may not be significant in practice for
planar impropers, in that the important thing is that the central
atom must be in the 3rd place, and both orderings are consistent
there. It might be a more serious problem if anyone had non-planar
impropers, such as for a united-atom ff, since it could reverse
the direction of force for the central atom. However non-planar
impropers aren't implemented.. if they were, then this particular
disordering might still be ok because the wrapped order of the
peripheral atoms is still the same for both: CD1-CD2-CG.

Bill

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