While we're on the subject...
We're writing a check into tleap that will alert the user if any torsion parameter is being overwritten by a previous declaration. I suppose we could add that for all the params.
We're thinking to only have them print at higher verbosity levels, but we defer to your preference.
:-) Lachele
Dr. B. Lachele Foley
Complex Carbohydrate Research Center
The University of Georgia
Athens, GA USA
lfoley.uga.edu
http://glycam.ccrc.uga.edu
________________________________________
From: B. Lachele Foley [lfoley.uga.edu]
Sent: Monday, May 02, 2011 3:36 PM
To: amber-developers.ambermd.org
Subject: [AMBER-Developers] Improper rules
In another email, DAC said:
> The criterion is this: the central atom is always at position 3; positions 1,2
> and 4 are filled in alphabetical order by atom type.
I agree wholeheartedly that this sounds like a perfectly reasonable rule. However, I do not believe that any of the leaps are following it. If it isn't being followed, what to do? Do we ditch the rule? Suddenly begin to enforce (seems dangerous)?
Briefly:
* In a file, we have an atom type N bound to types H, C and CG. So, that should match C-CG-N-H according to the rule above.
* In the (new, modified) version of GLYCAM that we've been playing with, the relevant improper torsion list looks like this:
X -X -N -CG 1.5 180. 2.
(that is, we removed the X-X-N-H entry)
* Both sleap and tleap match our atoms to the line "X -X -N -CG" in Glycam06. They both also match to the line "X -X -N -H" in parm99. We know the match is to X-X-N-H in parm99 because we tested by changing the value to "666" (no meaning intended there :-).
* When tleap writes the prmtop, the value assigned is based on the value present in the last parameter file loaded. That is, if we load parm99 last, we get 1.0 (or 666 in testing). If we load glycam06 last, we get 1.5.
* When sleap writes a top file (or as it used to; I'm sure Wei is working on this), it assigns both values.
When was the last time this alphabetical-order rule was enforced? We are leaning heavily toward getting rid of wild cards altogether.
To test this, it's simplest to use: "m = sequence { ROH 0YB }" and load whatever parameters you want to test. Note that's a zero and not an oh in 0YB, but it is an oh in ROH.
:-) Lachele
Dr. B. Lachele Foley
Complex Carbohydrate Research Center
The University of Georgia
Athens, GA USA
lfoley.uga.edu
http://glycam.ccrc.uga.edu
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Mon May 02 2011 - 18:00:02 PDT
