For very fine control over which residues get refined, I've added the
following to my .coot:
(add-key-binding "Add click to ref list" "a"
(lambda ()
(let ((imol-map (imol-refinement-map)))
(if (valid-map-molecule? imol-map)
(user-defined-click 1 add-to-reflist)
(show-select-map-dialog)))))
(define (add-to-reflist cl)
(format #t "clicked residue ~s~%" cl)
(let ((c (car cl)))
(format #t "cdr c is ~s~%" (list-tail c 2))
(format #t "head is ~s~%" (list-head (list-tail c 2) 3))
(set! reflist (append reflist (list (list-head (list-tail c 2)
3))))
(format #t "reflist is now ~s~%" reflist)
(user-defined-click 1 add-to-reflist)))
(add-key-binding "Refine cur list" "z"
(lambda ()
(begin
(clear-pending-picks)
(format #t "refining ~s~%" reflist)
(refine-residues refmol reflist)
(format #t "done refineing")
(set! reflist '()))))
With these functions, hitting the 'a' key initiates a mode where each
residue you click on is added to a list. When you then hit 'z' the
residues in the list are real-space-refined and the list is cleared.
You can use this to click on the metal and its coordinating residues and
refine only that selection. The only problem is that the cross-hair
picking cursor changes back to the regular arrow cursor after the first
click. Residue selection still works, you just have to believe. I
think Paul said he might do something that makes it possible to avoid
this problem in the future...
Thanks,
Joel
-----Original Message-----
From: Mailing list for users of COOT Crystallographic Software
[mailto:[email protected]] On Behalf Of Phil Evans
Sent: Thursday, November 18, 2010 12:45 PM
To: [email protected]
Subject: Re: [COOT] Colour by chain
On 16 Nov 2010, at 13:22, Paul Emsley wrote:
> Dear Phil,
>
> Sorry for the delay, I am only just back from somewhat extensive
travels.
>
> On 01/11/10 12:26, Phil Evans wrote:
>> If I "colour by chain" (my preferred option), metal ions disappear
>
> Thanks for the notification of this bug. It is fixed now, rev 3245.
thanks
>
>> also is there any way of refining a zone around a metal ion without
>> the sidechains vanishing into the metal density? I know the
>> refine-by-sphere ("R") option works better, but then I have no
>> control over what I'm refining
>>
>
> Sphere refine is simply a thin layer around the refine-residues
> function which takes a list of residue specs - that gives you enough
> control, I hope. (You can also trivially change the "radius" in
> sphere refine.)
>
> Paul.
How do you change the radius?
I had a long chat to Bernhard about this. My feeling is that in order to
avoid "user consternation" ((c) George Sheldrick), RS refinement in Coot
should _never_ move something into space already occupied by another
part of the same molecule (zero occupancy excepted). I realise there are
implementation difficulties, but I do think this is a major problem.
Phil
