Re: [modeller_usage] model.pick_atoms problem

[Modeller Caretaker <modeller-care@ucsf.edu>]

Alex Brown wrote:
> This may, or may not, be a problem, but I have found that using the 
> selection_segment argument ‘FIRST:A’ , ‘LAST:A’ in model.pick_atoms 
> doesn’t appear to work properly. In fact, the whole molecule (all 
> chains) is selected (i.e. the argument acts as if it were ‘FIRST:’, 
> ‘LAST:’). Correct selection does occur if discrete, numbered residues 
> are used, as in ‘1:A’, ‘100:A’, selecting atoms in a segment from 
> residues 1 to 100 in chain A.

The documentation is perhaps a little misleading; you can't use FIRST or 
LAST for selection_segment (it's only valid in an alignment file, or for 
model_segment when reading a PDB file). If the residue:chain is not 
found (as will be the case if you use FIRST or LAST) Modeller defaults 
to an all-atom selection.

The pick_atoms command will be replaced with a more flexible selection 
method in the next Modeller release.

	Ben Webb, Modeller Caretaker
--
modeller-care@ucsf.edu             http://www.salilab.org/modeller/
Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage