Sang,
They are always different. But depending on your data/parameter ratio,
you may be better of assuming they are similar (with NCS restraints) or
even identical (with strict NCS). Ask to a friendly crystallographer
around you when employing NCS is good for you. Crystallographers with
high resolution syndrome are not recommended to use NCS. Certain side
effects may include decreased R-Rfree spreads, and local mismatches
between electron density and model.
Engin
P.S. Gerard Kleywegt's papers from 1996 on NCS and refinement practices
might be a good place to start reading.
Sang Hoon Joo wrote:
I am refining my crystal structure in which I have two identical
chains in one asymmetric unit.
Space group is H32 and each chain yields me a biological trimer as expected.
The problem is, do I have to assume they are identical, or they are
really different.
After each cycle of refinement, if I try to align two molecules I get
~ 0.17 RMSD.
