If you view the value of the density as the number of electrons per
cubic Angstrom relative to the local average it is both accurate and
precise. This framing avoids the most serious problems with a measure
based on the rms, such as the problem that the "rms" depends on the
particular volume of space calculated (how much of the bulk solvent
region is included) and what stage of refinement is the project (late in
refinement the difference map will have a much smaller rmsd and the same
missing water molecule will have a much taller peak.).
When a map is viewed at a particular contour based on e/A^3 a
missing atom will tend to have the same difference map peak height in
all stages, while the height will vary all over the place when the
contour is based rmsd. When I watch people model building I see many
scrolling the contour level up and down without regard for the numeric
value of the level, apparently just choosing a value that makes the
peaks appear the way they prefer the peaks to appear. This way of
choosing a contour level seems dangerous to me as it seems to bias the
appearance of the map towards the modeler's expectation.
Dale Tronrud
On 5/2/2022 11:51 AM, Tim Gruene wrote:
Hi Ian,
thanks - this makes it clear(er) to me. The unit [e/A^3] suggested an
accuracy to me that isn't really there, while rmsd felt more what it
really is (maybe only due to habit and training).
Cheers,
Tim
On Mon, 2 May 2022 15:27:13 +0100
Ian Tickle <ianj...@gmail.com> wrote:
Hi Tim
I would say that it's not the displayed map density value in whatever
units that's arbitrary: it's completely defined as the true density
on an absolute scale minus the F000 contribution ('b' below) and
optionally divided by the RMS. It's just that we don't have a good
estimate of F000 and as I said it's your choice of contour level
that's arbitrary.
The Buster map is scaled to the model so is on an absolute scale. We
can write the linear transformation:
rho[map] = a.(rho[true] - b)
where the scale factor a = 1 so that rho[map] is on the same absolute
scale as rho[true] (i.e. differences in rho[true] are equal to
differences in rho[map]), and b is the F000 contribution.
Cheers
-- Ian
On Mon, 2 May 2022 at 15:04, Tim Gruene <tim.gru...@univie.ac.at>
wrote:
Hi Paul,
I'd rather you comment on the unit "e/A^3" displayed at the top of
my Coot canvas when I change the contour level of a map. Is this
arbitrary?
I noticed that the Buster Wiki states that their FWT is on an
absolute scale. Would there be a paper or url to advance my rusty
habits?
Cheers,
Tim
On Mon, 2 May 2022 13:52:05 +0100 Paul Emsley
<pems...@mrc-lmb.cam.ac.uk> wrote:
On 02/05/2022 12:44, John R Helliwell wrote:
I was intrigued by Paul’s 5sigma Fo-Fc default cut off, rather
than 3.
It's 5.6 now :-)
########################################################################
To unsubscribe from the COOT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1
This message was issued to members of www.jiscmail.ac.uk/COOT, a
mailing list hosted by www.jiscmail.ac.uk, terms & conditions are
available at https://www.jiscmail.ac.uk/policyandsecurity/
--
--
Tim Gruene
Head of the Centre for X-ray Structure Analysis
Faculty of Chemistry
University of Vienna
Phone: +43-1-4277-70202
GPG Key ID = A46BEE1A
########################################################################
To unsubscribe from the COOT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1
This message was issued to members of www.jiscmail.ac.uk/COOT, a
mailing list hosted by www.jiscmail.ac.uk, terms & conditions are
available at https://www.jiscmail.ac.uk/policyandsecurity/
########################################################################
To unsubscribe from the COOT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1
This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list
hosted by www.jiscmail.ac.uk, terms & conditions are available at
https://www.jiscmail.ac.uk/policyandsecurity/
