Hi Pavel and Joe, I carried out a test to see if (1) excluding a super-free set completely from automated structure solution affects the results and (2) if the super-free R that results is different from the standard free R obtained with all data and excluding the free set only from refinement. Here are the results: A = structure B, C and D are for the run with a super-free set excluded from all calculations. B and C are work and free R for the reflections included in the calculation. D is the R for the super-free set. E and F are work and free R for the same datasets, run in the same way, but without any super-free set. A B C D E F ---super-free-run---- --standard run-- structure work free super-free work free 1029B 0.31 0.38 0.41 0.29 0.36 1038B 0.22 0.26 0.26 0.24 0.28 1071B 0.28 0.33 0.35 0.26 0.30 rab3a-sad 0.31 0.35 0.38 0.29 0.34 rnase-p 0.28 0.32 0.31 0.28 0.29 calmodulin 0.26 0.27 0.27 0.27 0.29 cobd-sad 0.24 0.26 0.26 0.24 0.28 insulin 0.27 0.30 0.29 0.25 0.26 nsf-n 0.23 0.26 0.26 0.21 0.24 ut-synthase 0.24 0.29 0.28 0.24 0.28 mean 0.26 0.30 0.31 0.26 0.29 Conclusions: (1) super-free is very nearly the same as free in the matched set. (2) leaving out super-free data does not make a lot of difference in these actual cases. (Perhaps it might in worse data sets) (3) so it would be ok to take out super-free set, but it is not necessary. All the best, Tom T
Hi Pavel and Joe,
I'm glad you put that option in, Pavel. However for model-building it is not so straightforward. Normally we are building into density-modified maps. Density modification works poorly when a significant set of reflections is excluded, so this becomes impractical. We could exclude those reflections in the map only, but they will have been used in density modification so they are no longer truly free. I don't have a good solution for that.
I am guessing that for model-building, the free R is hardly affected, compared to refinement against the map, though I haven't tested this.
All the best, Tom T
Hi Joe, hi Tom,
until recently, all reflections (work and test) were used in map calculation. When I added a real-space refinement option to phenix.refine I spent a day thinking hard about why Rfree was almost always equal to Rwork after real-space refinement, until I realized that I have to exclude test reflections from map calculation, and that fixed the problem.
I think it depends on the task:
- for things like real-space refinement (where the map is heavily used) or automated model building, the test reflections have to be excluded from map calculation;
- for catching a tiny detail (looking at weak ligand density = slight map use) or fixing a few side chains, all reflections should be used, since 10% of data put aside may sometime significantly affect map quality.
Pavel.
On 9/30/09 11:51 AM, Tom Terwilliger wrote:
Hi Joe, In PHENIX R-free reflections are included in your maps. All the best, Tom T
On Sep 30, 2009, at 12:47 PM, Joe Krahn wrote:
Does PHENIX exclude R-free reflections when computing maps? IMHO, it is bad to use test reflections in maps, or anywhere else other than computing R-free, but opinions vary. If some people still want test reflections in maps, maybe it could be an option.
Thanks, Joe Krahn
phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb