Using G instead of Gr is part of the new PDB3 standard. If I
remember correctly, a DNA G is now called DG, and I think CNS still uses
GUA. Regardless, it sounds like your distribution of Coot may still be
using the old-style monomer libraries. The status message you pasted
below mentions UR.cif, which shouldn't exist in the new library. It
should be U.cif instead. Maybe the new library didn't make it into the
OS X build of Coot? If I unzip the stand-alone distributable of Coot
0.6.2
(http://psbmini.ucsc.edu/~wgscott/coot/Coot-0.6.2_10.6_and_10.7.pkg.zip
<http://psbmini.ucsc.edu/%7Ewgscott/coot/Coot-0.6.2_10.6_and_10.7.pkg.zip>),
it looks like it still contains the old-style monomer library.
You could try pointing $COOT_REFMAC_LIB_DIR at the CCP4 6.2 monomer
library (export COOT_REFMAC_LIB_DIR=//path/to/CCP4//lib). If that
doesn't work, you could also try downloading a Linux build of Coot (such
as
http://lmb.bioch.ox.ac.uk/coot/software/binaries/releases/coot-0.6.2-binary-Linux-i686-ubuntu-10.04.2-python-gtk2.tar.gz)
and un-tarring it. Then point $COOT_REFMAC_LIB_DIR at the Linux monomer
library (export
COOT_REFMAC_LIB_DIR=//path/to/Linux/coot//share/coot/lib). One of these
will hopefully fix the problem.
On a related note, it appears that the 64-bit Ubuntu Python build
of Coot also includes the old monomer library. At the very least, when
I un-tarred
http://lmb.bioch.ox.ac.uk/coot/software/binaries/releases/coot-0.6.2-binary-Linux-x86_64-ubuntu-10.04.2-python-gtk2.tar.gz,
it looks like it contains the old monomer library. This should mean
that 64-bit Ubuntu users would experience the same problem as Francis
and Huw. Paul, is this a bug or something intentional (or am I seeing
things wrong)?
Huw, I checked the B-factors under WinCoot using the new-style
monomer library, and I'm seeing the same B-factors as you. I'm not sure
if the large differences between the backbone and base are intentional
or not. Paul, do you happen to know where those numbers are from?
- Kevin
