Nathaniel Echols wrote:
> On Tue, Apr 28, 2009 at 3:52 PM, Oliver Clarke <[log in to unmask]
> <mailto:[log in to unmask]>> wrote:
>
> Is it possible in coot to script a way to colour residues by their
> position on the ramachandran plot? This would be really handy,
> particularly for large structures, in order to visually see where
> problem areas of the structure are. I'm thinking of something like
> alpha=green, beta=blue, left=yellow, and outliers=red. On a similar
> note, would it be possible to map the output of the rotamer analysis
> and density fit tools onto the structure by colouring each residue
> according to rotamer-likelihood or real-space CC?
>
>
> I have a more general version of this question: is there a function in
> the scripting interface that would permit atom- or residue-specific
> manipulations, of color or other properties? I would also find what
> Oliver describes useful, but I'm limited to doing this in PyMOL right now.
There is as yet no "colour by user property" (I would like to add this
at some stage - and map those values to a surface). Anyway, for now, we
can fiddle with pseudo B-factors. Something like this?
http://strucbio.biologie.uni-konstanz.de/ccp4wiki/index.php/COOT#Example_Scheme_Script_7:_Applying_arbitrary_value_to_.22B.22_factor_column
but where the values are computed from Ramachandran probability. That
should be quite tractable.
To answer Nat's question:
Atom properties can be set using the functions
set_atom_attribute()
and
set_atom_string_attribute()
You cannot set the colour directly.
(See the test suite for example usage.)
Paul.
|