A few comments -
you dont give your cell dimensions but if they are more or less the same
for the P212121 and P43 21 2 I dont see how a 43 set of absences can
turn into a 21 set ..
However if there is non-crystallographic translation, then absences an
mislead.. Is that true for you -
Tw
Thank you, Frank, for pointing this out
Here is the link to that presentation again:
www.ysbl.york.ac.uk/refmac/Presentations/Refmac_February.ppt
Garib
On 24 Apr 2010, at 12:05, Frank von Delft wrote:
Hi Garib, the link you sent doesn't work from here,
phx.
On 24/04/2010 00:17, Garib Murshu
I think your procedure is good with current technology. At early
stages twin refinement may give misleading results.
An intuitive resoning for low R factor would be: Twin is summation of
intensities. As you sum intensities two things happen: 1) distribution
of intensities become more symmet
Hello Garib,
I was using the ouput from a refmac run for the next round and using what I
thought were the original fobs .
So I guess as you said at each round I was using the de-twinned result from
the previous round.
I will repeat these refinements with the original scala output mtz to get a
more
Hari,
What twin tests have you run? Results? If your data really is
P43212 and you drop to P212121 you will still have the additional
two-fold operator in your data. An operator is a mathematical operator,
which could be crystallographic or twin.
I never refine a structure with twin
And another (not so) minor point: Defferences between R/Rfree with
twin on and off suggest that there is strong correlation between twin
and NCS. For this cases better dealt with if you use twin refinement
with sufficiently strong NCS restraints. In new refmac (www.ysbl.york.ac.uk/refmac/
You may be using output reflection data from the previous cycle of
refinement for the next session. After twin refinement output Fobs
(confusingly) may be detwinned. You can try to use original reflection
data file (e.g. after scala/truncate/freerflag) and for refinement
and it may clarif
I am refining a twinned dataset in possible spacegroup P212121 . Pointless
thinks it is P43212 , but based on reading this posting (
http://www.phenix-online.org/pipermail/phenixbb/2007-September/000501.html)
I think it is P212121.
The starting R/Rfree after molecular replacement ( single site mu
