[ccp4bb]: bulk solvent corrected fo-fc map [continued...]

[Mark DePristo <mdepristo@cryst.bioc.cam.ac.uk>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***

I've continued to try to solve the Fo-Fc problems with refmac5.  Here 
is an update:

The bulk solvent corrected maps in created by fft in xfit from Fo and 
Fc are extremely nice, as you can see at:

	http://raven.bioc.cam.ac.uk/~mdepristo/refmac/xfit-fo-fc.jpg

It is unambiguous which atoms are incorrectly placed in this map.  The 
MTZ file was converted using the following commands:

mtz2various hklin refine-0-12.mtz hklout refine-0-12.phs << eof
LABIN FP=FP FC=FC PHIB=PHIC
OUTPUT USER '(3I4,x,F7.2,3x,F7.2,3x,F7.2)'
END
eof

which dumps the H K L FP FC PHIC.  The phs file is available at:

	http://raven.bioc.cam.ac.uk/~mdepristo/refmac/refine-0-12.phs

However, the ccp4 fft version of Fo-Fc produced with DELFWT, PHDELFWT 
is clearly not as nice:

	http://raven.bioc.cam.ac.uk/~mdepristo/refmac/refmac-fo-fc.jpg

This map was generated with ccp4i's interface to fft, using a simple 
map with F1=DELFWT and PHI=PHDELFWT, as you can see at:

	http://raven.bioc.cam.ac.uk/~mdepristo/refmac/ccp4i-fft-fo-fc.jpg

then converted to fsfour format with map2fs.  The MTZ was produced by 
refmac5 (version 5.1.24) with the bulk solvent correct enabled with 
default values.  The refinement log is:

	http://raven.bioc.cam.ac.uk/~mdepristo/refmac/refine-0-12.log

and the produced MTZ is:

	http://raven.bioc.cam.ac.uk/~mdepristo/refmac/refine-0-12.mtz

and the output PDB file:

	http://raven.bioc.cam.ac.uk/~mdepristo/refmac/refine-0-12.pdb

I've tried several suggestions from CCP4bb readers:

	(1) I've done the fft of the DELFWT, PHDELFWT directly in xfit, and 
the resulting map is of poor quality.  This suggests that the poor map 
quality is caused directly by the DELFWT, PHDELFWT values output by 
refmac5, and not some strange effect of ccp4's fft.

	(2) Ian tickle tells me that refmac 5.1.29 produces much better maps, 
but I've been unable to download refmac (the refmac ftp site is empty).

	(3) The 2Fo - Fc map created using FWT and PHFWT is also poor.

	(4) The ccp4 fft of Fo - Fc directly (not with DELFWT) produces a map 
much nicer than DELFWT/PHDELFWT, but worse than the bulk solvent 
corrected xfit map.

		http://raven.bioc.cam.ac.uk/~mdepristo/refmac/fft-fo-fc.jpg

I've concluded that the Fo and Fc values produced by refmac are fine, 
as the xfit fo-fc map is very nice and the fft fo-fc is ok but lacks 
the solvent correction of xfit.  Something is going horribly wrong with 
the refmac DELFWT, FWT and PHDELFWT, PHFWT calculations.  Perhaps I'm 
not calculating the map correctly from DELFWT, but I think I'm doing it 
correctly.

Thanks for everyone's help so far!

Mark

Mark DePristo
Ph.D. Candidate
Dept. of Biochemistry
Cambridge University
mdepristo@cryst.bioc.cam.ac.uk
http://www-cryst.bioc.cam.ac.uk/~mdepristo/