Phil, I think the PDB documentation in this area (such as it is) is unclear, and it's not easy to fathom what was the original intent. The first line you refer to:
REMARK 3 NUMBER OF REFLECTIONS : appears in the section with the sub-heading: REMARK 3 DATA USED IN REFINEMENT. Arguably, and this was my initial impression, the test set is not 'used' in refinement, i.e. it has no effect on the refined parameters (except possibly indirectly via adjustment of the weights). So on that basis, Refmac is correct, the number of reflections 'used' should *exclude* the test set. However the second line you refer to: REMARK 3 FREE R VALUE TEST SET COUNT : appears in the next section with the other cross-validation info, with the sub-heading: REMARK 3 FIT TO DATA USED IN REFINEMENT. and seems to contradict this, since its use of the word 'used' (i.e. in the wider sense of the reflections merely being read in by the refinement program and passing the various rejection tests) clearly does imply that the test set is *included* in the count of 'used' reflections. So it all comes down to what is meant by 'used'. Cheers -- Ian On Wed, May 19, 2010 at 9:02 PM, Phil Jeffrey <pjeff...@princeton.edu> wrote: > Compare these two lines from phenix.refine: > REMARK 3 NUMBER OF REFLECTIONS : 46001 > REMARK 3 FREE R VALUE TEST SET COUNT : 2339 > > with those from refmac, ostensibly using the same data and start pdb: > REMARK 3 NUMBER OF REFLECTIONS : 43672 > REMARK 3 FREE R VALUE TEST SET COUNT : 2339 > > > I know there are 46011 reflections with |F|>0 in the files I used. > phenix.refine removes 10 of these as outliers. The 46001 remaining reported > in REMARK 3 *include* the test set. > > With REFMAC, 43672+2339=46011 so it appears that Refmac reports just the > *working* set count in that first line, excluding the test set. > > Is this is a bug with one program or the other, or a bug in the PDB > definition of REMARK 3 ? > http://www.wwpdb.org/documentation/format23/remark3.html > > This appears to be a source of inconsistency. > > phenix.refine 1.6-289 > refmac5 5.4.0077 (I'm apparently a Luddite) > > Phil Jeffrey > Princeton >
