[phenixbb] statistics for highest shell in phenix.refine
Pavel Afonine
PAfonine at lbl.gov
Wed Sep 5 15:03:14 PDT 2007
Hi Michael,
phenix.refine reports the Rwork- and R-free factor statistics in bins
(as well as completeness, fom, phase error estimates, number of
reflections, etc). Look in you .log file for a table like this:
|-----------------------------------------------------------------------------|
| Bin Resolution Compl. No. Refl. R-factors
Targets |
|number range work test work test
work test|
| 1: 24.6802 - 3.7756 0.97 1800 135 0.2554 0.3150 6.0946
6.2202|
| 2: 3.7756 - 2.9985 0.99 1845 140 0.2466 0.2738 6.1306
6.4268|
| 3: 2.9985 - 2.6200 0.99 1819 156 0.2738 0.3065 5.9144
6.1387|
| 4: 2.6200 - 2.3806 0.99 1820 143 0.2662 0.3266 5.7197
5.8192|
| 5: 2.3806 - 2.2101 0.99 1813 138 0.2515 0.3138 5.5753
5.6897|
| 6: 2.2101 - 2.0799 0.99 1844 135 0.2518 0.2708 5.4776
5.5673|
| 7: 2.0799 - 1.9758 0.99 1818 140 0.2512 0.3325 5.3948
5.728|
| 8: 1.9758 - 1.8898 0.99 1806 144 0.2652 0.3208 5.3259
5.5097|
| 9: 1.8898 - 1.8171 0.98 1833 137 0.2450 0.2865 5.1955
5.2526|
| 10: 1.8171 - 1.7544 0.98 1765 165 0.2660 0.3250 5.0805
5.2769|
| 11: 1.7544 - 1.6995 0.99 1872 128 0.2538 0.3600 4.9928
5.2402|
| 12: 1.6995 - 1.6510 0.98 1770 152 0.2550 0.2989 4.8696
5.0337|
| 13: 1.6510 - 1.6075 0.96 1798 153 0.2609 0.2944 4.7976
5.0056|
| 14: 1.6075 - 1.5683 0.95 1708 137 0.2554 0.3106 4.7096
4.9238|
| 15: 1.5683 - 1.5327 0.90 1686 145 0.2450 0.3115 4.6753
4.8214|
| 16: 1.5327 - 1.5000 0.86 1590 110 0.2452 0.3547 4.609
4.8343|
|-----------------------------------------------------------------------------|
In this particular example above the numbers you are interested in are
located in the line # 16: Rwork=0.2452 and Rfree=0.3547.
A few things to keep in mind:
1) phenix.refine prints this table several times: before, during and
after refinement, so make sure you're looking the latest one (printed
just before refinement finishes);
2) If you're using the latest version of phenix.refine (from PHENIX or
CCI Apps), then the PDB file header contains similar statistics in bins
so you don't need to look at long .log file.
Please let me know if you have any questions.
Pavel.
michael hothorn wrote:
> Dear all,
>
> For the crystallographic table in my manuscript, can I define a highest
> shell in phenix.refine and have it output the refinement statistics for
> this range? If not, how can I calculate the highest shell statistics
> using the same parameters as used for the overall refinment?
>
> Thanks
> Michael
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://www.phenix-online.org/mailman/listinfo/phenixbb
More information about the phenixbb
mailing list