Dear Michael,

     In sftools you can generate a set of FreeR flags on only a subset of
the reflections.  If you have a combined dataset (high and low relsoution), then
select the high res refls, construct the FreeR set (unselected refls should not 
be changed) and write out.


sftools
READ MTZFILE
SELECT RESOL < Res_Limit
RFREE 0.05 SHELL
SELECT ALL
WRITE ANotherFile.mtz
QUIT


Only think to note is that the generated column will be called RFREE. Keep an 
eye on
how many shells are used, might have to make them thicker. Finally to select 
high res 
refls  use < not > .


Adam














    





_______________________________________________________________________
Good Morning,

I am solving a structure with 6-fold NCS and consequently used sftools 
to select a 5% test set of reflections in thin resolution shells to 
monitor Rfree.
I have reprocessed the same dataset to higher resolution to improve the 
quality of the electron density maps and now need to extend the test 
set.
There appears to be no direct method in sftools of keeping the same 
test set while adding an additional set in the same proportion to 
higher resolution, also in thin shells.
The only way I can see this might be done is to:
1. SELECT RESOL > to give the high resolution shell of reflections
2. Generate the test set from the high resolution shell using RFREE 
0.05 SHELL.
3. SELECT RESOL > to produce a reflection file with the same resolution 
range as the old dataset.
4. READ [old dataset] COL RFREE.
5. SELECT COL RFREE=0.
6. SELECT ALL to include all reflections.
7. WRITE New_Dataset.mtz.

The result should be a higher resolution dataset, with 5% test set of 
reflections in thin resolution shells, keeping the same test 
reflections as the old set.
Will this work, or is there an alternative reliable procedure?

Best wishes,

Michael

Michael Roberts
Crysalin Limited
Cherwell Innovation Centre
77 Heyford Park
Upper Heyford
Oxfordshire OX25 5HD
Phone: 01869 238030

Reply via email to