[phenixbb] SA

Irina PROKHOROVA prokhoro at igbmc.fr
Fri May 1 15:29:54 PDT 2015


Thanks, it worked but not the way I expected.

It definitely found problematic regions in the model but moved it not correctly. Should I adjust the parameters of SA? (I did torsion SA with a default settings and before MR and rigid-body by domains)

I am still improving the model since it has some clashes. There are also disordered regions and tiny movements of big domains. Is it neccessary to adjust the model as much as possible with rigid-body before SA?


Irina

________________________________
От: Pavel Afonine <pafonine at lbl.gov>
Отправлено: 30 апреля 2015 г. 15:30
Кому: Irina PROKHOROVA; phenixbb at phenix-online.org
Тема: Re: [phenixbb] SA

Hi Irina,

SA scope of parameters needs "refinement." prefix:

refinement.simulated_annealing {
    mode = every_macro_cycle second_and_before_last once *first first_half
}

Let me know if it doesn't do it!

Pavel

On 4/30/15 2:20 AM, Irina PROKHOROVA wrote:
Hello,

I am trying to run coordinate refinement 5 cycles with first cycle of simulated annealing (torsion) via command line. Here is a part of my params file that I adapted from my colleges:

refinement.main {
    number_of_macro_cycles = 5
    target = *ml mlhl ml_sad ls
    simulated_annealing_torsion = True
}

simulated_annealing {
    mode = every_macro_cycle second_and_before_last once *first first_half
}

refinement.refine {
    strategy = *individual_sites individual_sites_real_space rigid_body \
                individual_adp group_adp tls occupancies group_anomalous
    sites {
      individual = segid *A
      individual = segid *B
    }



As a result I get the Error
Unused parameter definitions:
  simulated_annealing.mode (file "refine.params", line 26)
Could you help me with the syntax, please. Should I include sites definition for simulated annealing?

Irina



_______________________________________________
phenixbb mailing list
phenixbb at phenix-online.org<mailto:phenixbb at phenix-online.org>
http://phenix-online.org/mailman/listinfo/phenixbb
Unsubscribe: phenixbb-leave at phenix-online.org<mailto:phenixbb-leave at phenix-online.org>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20150501/86fac066/attachment-0001.htm>


More information about the phenixbb mailing list