[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ccp4bb]: DM problem ... low resolution problem?

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

On Fri, 30 Jun 2000, Cusack Stephen wrote:

> dm: (CALCSCALE) - insufficient resolution: enter a SCALE card.
> Can anyone give advice to get round this please?

I think the obvious work-around would be to enter a SCALE card.

<aside:  how many of today's grad students understand where this use of
'card' arose?>

some density modification techniques, such as histogram
matching and Sayre's eqtn, are heavily dependent on the data being on the
'correct' scale so that in direct space the density values are accurate in
terms of electrons per cubic Angstrom.
typically you will approximate the 'correct' scale in
TRUNCATE by Wilson methods, but these are not accurate at low
resolution (~ 4 or 5 A is the cutoff). DM has some internal
routine to calculate the scale, but obviously it is not comfortable doing
so at your data resolution.

the syntax for SCALE is covered in the DM documentation.  As to how you
might come up with a reasonable value, I don't have any intelligent
solutions.  12-fold NCS averaging is a pretty strong constraint (assuming
you got it right) so you could run with that and solvent flattening,
neither of which depend on true scale, and turn off the histogram matching
by putting a NOHIST keyword in the MODE line.  At your resolution you will
certainly not be running Sayre's eqtn.


"I'm not going to be anybody's puppet, particularly not my own."
       - Zaphod Beeblebrox
                        David J. Schuller
                        modern man in a post-modern world
                        University of California-Irvine