[buster-discuss] OMIT maps in BUSTER

[Clemens Vonrhein]

Dear Guillermo,

On Mon, Sep 26, 2016 at 10:08:44AM +0000, Guillermo Montoya wrote:
> I´ve been searching in the web docs how to prepare an OMIT map in BUSTER
> but I could not find the command

What kind of OMIT map are you referring to? Of course, you can just
omit a part of the model and re-run BUSTER in order to get a map.

If you are thinking of either a composite omit map or some kind of
e.g. simulated-annealing omit map, there is no dedicated feature in
BUSTER for doing this.

However, we never found it a feature we were desperately missing:
BUSTER maps seem to have very little model bias (we could show dozens
of examples where model errors show up immediately in a BUSTER map -
even before any refinement). You could maybe test this yourself by
mis-building a small part of your model on purpose and see what
happens during refinement with BUSTER.

Cheers

Clemens

PS: if that is triggered by a request from a referee, it might be that
    someone is still thinking of the good (?) old X-PLOR/PROLSQ days
    when everyone had massive model bias issues with our least-square
    formulation. Not sure if composite/SA-OMIT maps are any longer
    useful or needed to prove a point (but: old habits die hard). I
    might be wrong though.

-- 

*--------------------------------------------------------------
* Clemens Vonrhein, Ph.D.     vonrhein AT GlobalPhasing DOT com
* Global Phasing Ltd., Sheraton House, Castle Park 
* Cambridge CB3 0AX, UK                   www.globalphasing.com
*--------------------------------------------------------------