Re: Masking

[<naismith@st-andrews.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    ***

Since I started this arguement let me supply my own definitions and 
perhaps supply a little more clarity




We like DM's solvent flattening mask however, I think it could be improved.
It is after all starting from a very poor map.
In our hands automasking of NCS molecules fails.
(even in another example where we have quite a reasonable starting point).
As I said the multimer approach to pure ncs whereby a sphere is drawn at 
the centroid works pretty well in several examples. So an NCS mask we do not 
need.

My question is how to improve the orginal DM solvent mask.

My thoughts were if we apply the ncs matrix to the cell and 
superimpose and average.
As more than one person pointed out this would avergae non-identical things.
True so they would hopefully cancel or at least stay noisey.

However, for the asu it should average monomer A onto monomer B and as it 
is pure NCS, monomer B onto A . Thus monomer A and monomer B should rise 
from the noise at least a little. Hopefully generating a better protein 
solvent boundary and thus a better intial solvent flattening mask.

This is my definition of maskless averaging. We make no assumption about 
the extent of monomer A or monomer B. Of course it is a one shot deal, we 
do not recycle phases etc.

best
Jim

James H. Naismith                | Research mailto:naismith@st-and.ac.uk
Centre for Biomolecular Sciences | Teaching mailto:jhn@st-and.ac.uk
BMS Laboratories                 | Office: 1334-463792 (24 hr)
The North Haugh                  | Lab   : 1334-463776
The University                   | Fax   : 1334-467229
St. Andrews                      | In UK     add  0 to start of number
Fife Scotland                    | Overseas  add 44 to start of number
U.K., KY16 9ST                   | www     http://speedy.st-and.ac.uk/