Dear Eleanor,
Thank you so much.
As you said, it is a problem that the number of reflections estimated by FreeR
set is different from an actual that of reflections because of the low
resolution discrepancy. Moreover, twin refinement includes a problem about some
overlapping of diffraction
Well - I am not sure whether this has anything to do with twin refinement
but resolution limits are often a bit iffy.
First the low resolution discrepancy. The Free R is generated to a
lower and higher resolution than any observation. The FreeR set is meant to
be complete for any possible index
When I use refmac restrained refinement with intensity based twin refinement,
resolution limit has been changed from 71.4519- 1.9996 to 47.392-1.986.
From Log file, before refinement, Scaling and SigmaA resln: 71.4519 1.9996,
seems to be correct.
But in the middle of log file, around Fom and
