[phenixbb] A few questions

Pavel Afonine PAfonine at lbl.gov
Wed Feb 6 11:48:18 PST 2008

Hi Chris,

thanks for the data. I'm looking at your parameters file (test2). In the 
strategy keyword:

strategy = *individual_sites rigid_body *individual_adp group_adp tls \
               individual_occupancies *group_occupancies group_anomalous

I see you are using group_occupancies which means that one occupancy per 
residue will be refined (unless you provide your own selections to 
define the groups). As I mentioned before,  group_occupancies does not 
use any constraints and it is not aware of alternative conformations. 
This is why in the example below the sum of occupancies is not 1.0. I 
think this is not what you want to do. I will add the constrains to 
group_occupancies option so you get equal occupancies within each 
conformer and the sum of occupancies computed for all conformers is 1.0.

Currently, if you choose individual_occupancies then you will get the 
constraints, that is the sum of occupancies for "523 SE AMSE A 32" and 
"530 SE BMSE A 32" will be one, but the occupancies for atoms within 
each conformer will be different, which is, as Paul correctly pointed 
out, not 100% right.

I will keep you posted about the progress.


> In that case, then the occ of alt conf's tends to refine to greater than 1:
> ATOM    523 SE  AMSE A  32      23.654  11.792   6.216  0.56 14.04          Se
> ATOM    530 SE  BMSE A  32      23.570  12.284   6.796  0.46 13.91          Se

More information about the phenixbb mailing list