Subject: Re: [nmr_sparky] overlay (ol)
From: Thomas Goddard
Date: Mar 30, 2010
Previous: 665
Hi Eiso,
I see. This is a bug. With your two spectra with axes 15N, 1H, 1H
Sparky will always match w1 - w1, w2 - w2 and w3 - w3. This is
the first matching rule I described before. The bug is that the overlay
doesnt work unless the two view windows have the same spectrum axes
(w1,w2,w3) displayed on the horizontal and vertical axes. The actual
axes displayed on the vertical and horizontal axes (which you can change
with xr, xx) isnt supposed to have any effect on axis matching or the
overlay. It in fact doesnt have any effect on axis matching in the
code. But then the code doesnt use the correct axes when trying to
draw the overlay. I dont think I can spend the time to fix this and
make a new Sparky release but it is good to document the problem.
Tom
Hmm, this doesnt quite correspond to what Im seeing.
these spectra I want to overlay are the same just with a bit
different processing params.
$ ucsfdata 3dnoesy.ucsf
axis w1 w2 w3
nucleus 15N 1H 1H
matrix size 256 512 1024
block size 8 16 32
upfield ppm 104.145 -1.727 4.773
downfield ppm 130.141 11.273 11.286
spectrum width Hz 2106.816 10395.982 5208.333
transmitter MHz 81.041 799.695 799.695
$ ucsfdata 3d.ucsf
axis w1 w2 w3
nucleus 15N 1H 1H
matrix size 256 512 1024
block size 8 16 32
upfield ppm 104.145 -1.727 4.773
downfield ppm 130.141 11.273 11.286
spectrum width Hz 2106.816 10395.982 5208.333
transmitter MHz 81.041 799.695 799.695
Im opening the spectra in the orientation x,y - w3,w2 (H-acq,
H-indirect)
and Iam seeing the overlay. Are you saying this shouldnt happen because
sparky cant match H(w2) and H(w3)?
Then if I use xr to view a peak from a different plane (e.g. w1,w3) the
overlay
disappears. But I can do remove and add overlay and then the overlay is
back
again. (But disappears again when I do xr again)
This doesnt seem to be related to whether axis can be matched or not....
kind regards,
Eiso
Tom Goddard wrote:
Hi Eiso,
Sparky does not use the displayed orientation of the spectrum axes in
the view window to decide how to pair the spectrum axes when
overlaying. Instead it first checks if the nucleus types (H, C, N) are
the same for the w1 axes in both spectra, for the w2 axes in both
spectra, and the w3 axes for both spectra. In that case it simply
pairs w1- w1, w2- w2, w3- w3 in the overlay. Otherwise it
tries to pair axes that have matching nucleus types. For example if
one spectrum had (w1 N, w2 C, w3 H) and you overlay another with (w1
C, w2 H, w3 N) then it will pair the axes with matching nuclei even
though they are in different order. If this mapping is not unique
because two nuclei a identical, for example, overlaying (w1 N, w2 H,
w3 H) onto (w1 H, w2 N, w3 H) then I believe Sparky just doesnt show
the overlay. The only way to get that case to work is to have the same
nuclei for axes w1, w2 and w3. So youd need to reorder the axes in
your spectrum file. So basically Sparky wont let you pair whichever
axes you want, it uses a few fixed rules -- unfortunate.
Tom
hi Tom,
I noticed that if you overlay 2 3d spectra, both spectra are only shown
in the orientation of the spectrum that was present during setting
the overlay.
Is this intentional? or a bug?
kind regards, Eiso