[phenixbb] refinement / alpha_beta

Pavel Afonine pafonine at lbl.gov
Tue Oct 5 22:08:17 PDT 2010


  Hi Kay,

the coordinate error is estimated from alpha/beta parameters, and 
alpha/beta parameters are estimated using test set of Fobs and Fcalc 
computed from your current model. So, if your model is poor (severely 
incomplete) then it obviously gets reflected on Fcalc and you get what 
you get.

I guess a way around would be to use an a priori knowledge about the 
missing structure (amount and location) and incorporate it into alpha 
and beta but this is not implemented.

I played with computing (as opposed to estimation) of alpha and beta 
(the parameters you refer to below) but it never worked good in my 
hands, although may be I didn't try hard.

I guess it is a nice project to spend a few weeks/a moths on but there 
is no quick solution right now, sorry.

Pavel.


On 10/5/10 11:30 AM, Kay Diederichs wrote:
> Dear Phenix developers,
>
> I'd like to refine (using phenix.refine) a _very_ incomplete model 
> (obtained by molecular replacement with one subunit of the complete 
> structure) against the native amplitudes, with the goal to obtain 
> meaningful phases and figures_of_merit which should serve to find 
> heavy atom sites in derivative datasets.
> If I "just refine" in phenix.refine, then the R-factors are in the 50s 
> and the program estimates the coordinate error to be very large. Thus 
> I would like to tell phenix.refine that the coordinates are actually 
> quite accurate, but that a lot of the model is simply missing.
> I found
>    alpha_beta
>       free_reflections_per_bin= 140
>       number_of_macromolecule_atoms_absent= 225
>       n_atoms_included= 0
>       bf_atoms_absent= 15.0
>       final_error= 0.0
>       absent_atom_type= "O"
>       method= *est calc
>       estimation_algorithm= *analytical iterative
>       verbose= -1
>       interpolation= True
>       fix_scale_for_calc_option= None
>       number_of_waters_absent= 613
>       sigmaa_estimator
>          kernel_width_free_reflections= 100
>          kernel_on_chebyshev_nodes= True
>          number_of_sampling_points= 20
>          number_of_chebyshev_terms= 10
>          use_sampling_sum_weights= True
> in the documentation but I have trouble understanding its meanings and 
> defaults - but I do think this should be used. Maybe I should switch 
> to method=calc and specify number_of_macromolecule_atoms_absent= as 
> well as bf_atoms_absent= ?
>
> Can anyone advise me, please?
>
> thank you,
>
> Kay
>
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20101005/e3e9c49f/attachment-0003.htm>


More information about the phenixbb mailing list