On Tue, 2005-08-09 at 14:42, Rasmus Fogh wrote:
> Dear Henry,
> 
> First some reasons:
> You clearly have a T1List where several T1's point to the same Resonance.
> This is illegal. Once upon a time neither Analysis nor teh API checked
> this properly, so it could be that the error got in with an old version of
> wither and is only now being found. If this cannot be the case somthing
> must have bypassed the API somehow, you might have edited teh file, or
> just possibly there could be a bug. If you have just merged Resonances
> that might be a place we could look - the merge code is necessary but in
> some ways rather fragile. Anyway, we would have to know what you have been
> doing, the versions you are using now and were using to write the files
> etc., to find out what happened. Some shourt summary would be nice, just
> for our curiosity.

I have been using Analysis 1.0.3 (haven't upgraded yet but will do once
this problem is sorted).
> 
I remember using the merge Resonance yesterday. I was assigning an
Asparagine sidechain NH. When I propagated the assigned 1H and 15H
dimensions of the Asparagine NH from the HSQC to the NOESY strip I
noticed that the 1H (direct) dimension for all Noesy peaks were assigned
to both 119AsnH and an unassigned resonance number (possibly resonance
[226] but I don't remember for sure). As I recall, I thought that by
merging these resonances in the resonance menu I would simplify matters
by having only one resonance for the 1H direct dimension of this
sidechain NH. This may be the source of the problem however I don't
understand how this would interfere with my T1 assignments unless the
resonance 226 was present in the HSQC when I use the Group Peaks
function in Data analysis -> Rate analysis.
> The Fix:
> Of course it is nicer to fix the problem and not to worry about its
> history. So, step by step:
> 
> Find the Resonance with serial=226, and note the _ID attribute. Let us
> call that _x. Now find the T1 XML elements elements that have the XML
> attribute 'resonance="_x"'. There should be a case where you have more
> than one of those within a single T1List.
Yes I had pairs of the same resonance for each T1list_ID. I have removed
one from each.
> Remove all but one of these.
> Make a note of the _ID of the T1 you are removing (let us call it "_y").
> Search through the XML file for any links to these objects (basically
> attributes that have _y as their value. I doubt there will be any, but
> there could be Peaks or PeakDims that are linked to your T1's. Remove the
> links (but not the objects that hold the links). 

I did a search for the T1 _ID="_y" in the entire Nmr.xml file after
removing them from the T1lists and found no matches.

However I did find matches for "_y" alone.
eg. for "_21257" I found several <measurements _IDREF="_21257"/>
associated with different Peak_IDs and I found a <t1s _IDREF="_21257"/>
Which of these and what part of these do I remove?

Henry

> Save the file and try to
> load it again.
> Good luck,
> 
> Rasmus
> 
> 
> ---------------------------------------------------------------------------
> Dr. Rasmus H. Fogh                  Email: [log in to unmask]
> Dept. of Biochemistry, University of Cambridge,
> 80 Tennis Court Road, Cambridge CB2 1GA, UK.     FAX (01223)766002
> 
> On Tue, 9 Aug 2005, Henry G. Hocking wrote:
> 
> > Hello,
> >
> > I received an error message this morning when trying to open up my
> > project.
> >
> > The pop up reads:
> >
> > Project invalid, please quit, fix and re-start: Non-unique key
> > <ccp.Nmr.Resonance [226]> for ccp.Nmr.T1List.measurements.
> >
> > and at the shell command line:
> >
> > last xml element read:
> > </_StorageUnit>
> >
> > parser state was:
> > checking  object validity
> >
> > ERROR loading xml ccp.Nmr ccp/Nmr.xml
> >
> > I'm not too sure what I would have done yesterday to cause this.
> > Where do I look for resonance [226] if I cannot open my project?
> >
> > Henry
> >