Re: [ccp4bb]: DM - NCS averaging

[Kevin Cowtan <cowtan@ysbl.york.ac.uk>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***

OK, this was a very interesting problem, and the log file gave the solution.

The problem is that the Spacegroup is P212121, but there is 2-fold NCS which 
is almost exactly crystallographic, and gives an approximation to P41212.

Because the NCS is pseudo-crystallographic, and your starting operator is 
exaclty crystallgraphic, the map is correlated over the whole unit cell. 
Therefore there is absolutely no way that local correlation-based automatic 
NCS mask generation can work. The mask you have covers a completely random 
portion of the unit cell.

The drop in correlation was due to the fact that you asked the program to 
make a new mask after 7 cycles, and it got something else completely random.

Since you NCS is so close to crystallographic, you are not getting any new 
information from it. So I suspect that NCS averaging is a waste of time in 
this case anyway.

Kevin