[phenixbb] sum(occ)<1?

[Pavel Afonine]

  Hi Ed,

just to make sure you are looking at the updated manual:
http://www.phenix-online.org/documentation/refinement.htm#anch20

Yes, you are right, ideally one would constrain occupancy between Qmin 
and Qmax. Answering your question: no, with the current version you 
cannot constrain the sum of occupancies to <1 (or to <Qmax), it is 
constrained to be equal to 1.

Special case: If you refine one occupancy per selected group of atoms, 
you can constrain it to be less than Qmax and larger than Qmin (1 and 0, 
by default).

It is in todo list, but it doesn't have high priority. Frankly, I don't 
know how to do this cleanly - minimization with inequality constraint 
(as opposed to what we have no - minimization with strict constraint). 
I'm sure there are right methods for this, but it might turn into a few 
months project to accomplish for unclear gain.

Pavel.


On 7/15/10 3:27 PM, Ed Pozharski wrote:
> Let's say there are two orientations of a ligand for which occupancies
> need to be refined.  By default, the sum of the two will be fixed at
> 1.0.  From quick look at the manual, the only way to have the sum(occ)
> at lower level is to reset main.occupancy_max.  However, this will
> probably affect all the atoms for which occupancies are refined, which
> is undesirable (while ligand may have total occupancy of<1, alternate
> conformers for protein residues should always add up to 1).
>
> Am I missing something or the only way to get this done is to refine
> twice - once only for ligand occupancy, and then for the rest of the
> model.
>
> And what if total occupancy of the ligand is unknown?  Is there a way to
> allow sum(occ) to be anywhere between 0 and 1?
>