Dear Takebe Katsuki,
Your statistics are not that bad. But first of all, you should not judge the resolution of your data set based on these R-values (Rwork or Rfree). With these R-values, you should not exceed the the values of 0.42 (R-value of a perfect model against the random data)! In general, the reflections from the higher resolution shells are weaker and have lower I over sigma. This is also the reason why the Rwork and Rfree values are higher for these shells. Another issue are the reflections from the low resolution. But, it is quite common to have your distribution.
Nevertheless, if you wanted to know how to best set the resolution, check the web sites of our program PAIREF:
https://pairef.fjfi.cvut.cz/ (Ooops, our certificate has expired, I will try to solve the issue immediately!)
The paper about the program is here: https://journals.iucr.org/m/issues/2020/04/00/mf5044/index.html
PAIREF is very easy to install. It also works with PHENIX.REFINE. However, paired refinement is also available somewhere in PHENIX (probably in some expert ... whatever).
We have also organized a webinar on how to work with PAIREF. You can check the video here: https://pairef.fjfi.cvut.cz/docs/pairef_poli_webinar/PAIREF_webinar_23Mar202...
I know, it may be a lot of information in one bunch. Feel free to ask off the list.
Best regards,
Petr
________________________________________
From: [email protected]
