Can you combine the PHIC and FOM from each of your solutions, then do
the Phase comparison in Clipper utilities checking for alternate
origins? That gives a good indication of what is going on.
Do you have more than one molecule in the asymm unit? You can finish up
with A_phaser matching one symm equiv of B_molrep and B_phaser matching
a different symm equivalent of A_molrep..
Eleanor
Kolstoe S.E. wrote:
Dear ccp4bb,
In searching for a mr solution I ran both phaser and molrep, with both
programs finding nice solutions in P21212 (cell 81.083 104.435
109.732 90.00 90.00 90.00, resolution 2.65A). For various reasons I
refined the phaser and molrep solutions separately, and both behaved
in exactly the same way giving R=0.24, Rfree=0.28 after adding
ligands, mutations etc. Today I decided to compare the two models
(which I figured were exactly the same) however found that the
asymmetric units did not superimpose. Normally this is just a case of
each mr program finding a different asymmetric unit in the same cell,
however this time there is a distinct clash between the two solutions
with about 50 atoms overlapping. After playing around with various
symmetry equivalents I have found that the difference between the
solutions is approximately a 30A translation along the x axis. As
there is no signs of any twinning, and both solutions behave in
exactly the same way in refinement, is this a case of phaser and
molrep using different origins or is there something more insidious
going on?
Thanks for any advice,
Simon
