Hi,
On Mon, Oct 25, 2010 at 08:36:18PM -0500, Artem Evdokimov wrote:
> As a very basic rule of thumb, Rfree-Rwork tends to be around Rmerge for the
> dataset for refinements that are not overfitted.
That got me thinking - and I went to check recent PDB entries
(2009-2010) with
0.0 < R_work < 1.0
0.0 < R_test < 1.0
0.0 < Rmerge < 1.0
(entries with differing values of Rmerge/Rsym were excluded for
simplicity).
That leaves 3026 entries. I attach two plots that might (or might
not?) show a relationship between Rfree-Rwork and Rmerge ... or maybe
they don't show it, I don't know.
-------------------------------------------------------------
Anyway, the interesting entries (at least for me) are the outliers in
those plots. There are some very odd values recorded in some entries
- e.g.
a) quite low Rmerge values:
Rmerge Rwork Rfree Rfree-Rwork Resolution
---------------------------------------------
0.0052 0.2140 0.2480 0.0340 1.90
0.0055 0.1980 0.2200 0.0220 1.80
0.0100 0.2070 0.2729 0.0659 3.00
0.0100 0.1510 0.1870 0.0360 1.85
0.0270 0.1476 0.2121 0.0645 1.60
b) very large Rmerge values:
Rmerge Rwork Rfree Rfree-Rwork Resolution
---------------------------------------------
0.9990 0.1815 0.2086 0.0271 1.80 <<< SG center, unpublished
0.8700 0.1708 0.2270 0.0562 1.96 <<< unpublished
0.7700 0.1870 0.2297 0.0428 1.56
0.7600 0.2380 0.2680 0.0300 2.50 <<< SG center, unpublished
0.7000 0.1700 0.2253 0.0553 1.71
0.6400 0.2179 0.2715 0.0536 2.75 <<< SG center, unpublished
The most disturbing to me is that of those with very large overall
Rmerge values, 3 come from structural genomics centers. Unfortunately,
these are all marked as 'unpublished', so I can't go back to a
paper or supplementary material to check (if there was just a typo in
the submission - or something else went wrong).
But an entry with 1.8A resolution, overall Rmerge of 0.99 and R/Rfree
of 0.1815/0.2086?
Cheers
Clemens
--
***************************************************************
* Clemens Vonrhein, Ph.D. vonrhein AT GlobalPhasing DOT com
*
* Global Phasing Ltd.
* Sheraton House, Castle Park
* Cambridge CB3 0AX, UK
*--------------------------------------------------------------
* BUSTER Development Group (http://www.globalphasing.com)
***************************************************************
|