Re: [ccp4bb]: DM

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

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

OK, there are several issues.

If your NCS is not proper, you CANNOT use a 10mer mask. The NCS will be
proper if the 2-fold axes intersect and are perpendicular to the 5-fole
axis. If it is not proper, you will need a 5mer mask, and the automask
routine will be give you this if you specify nmer 5 and make sure the
first non-identity ncs op is a 5-fold rather than 2-fold operator.

If you do have proper 10-fold symmetry, then you can use at your choice
a monmer, dimer, 5mer or 10mer mask. The choice depends on how good the
ncs axes are.

The advantage of using a 10mer mask is that if the map is very poor, the
operators will refine better because there is more info to go on. But,
if the NCS is not quite an exact 5-fold, then the averaging will blur
stuff a bit. (The matrix refinement is essentially locked in this case).

The advantage of using a monomer mask is that if the NCS is not exact,
then the matrices will refine individually. But the matrix refinement
might blow up.

If you can get a monomer mask, and the matrices don't blow up, then this
is probably the preferred approach. Otherwise, use the automask which
will generate the largest mask consistent with the first two ncs ops
supplied.

Kevin