Re: [Psidev-ms-dev] Chromatogram and spectrum terms in the CV

[Darren K. <dke...@ya...>]

Hi everyone,

I've added chromatogram support to ProteoWizard, including indexing.  
I've updated the tiny.pwiz.mzML example linked from the mzML page.

In addition to the CV terms mentioned by Matt, I think we should add a 
"dataProcessingRef" attribute to the <chromatogram> element in the spec, 
so we can encode where we are getting the chromatogram from (e.g. 
XCalibur directly vs. a post-processing calculation).

I haven't checked the schema, but it also needs to handle multiple 
<index> elements -- one to index into <spectrumList> and one to index 
into <chromatogramList>.  Currently <indexOffset> points to the first 
<index>, and we look for the multiple <index> elements.  If this bothers 
anyone, we might consider having either multiple <indexOffset> elements 
with name attributes, or put the <index> elements in an <indexList>.


Darren



Matthew Chambers wrote:
> Hi all,
>
> With the new developments in the schema to support chromatograms and 
> non-mass spectra, some corresponding changes to the CV should happen.
>
> The current "spectrum type" should be renamed to "mass spectrum type" 
> and a new "spectrum type" term should be defined that allows for 
> non-mass spectra:
>   
>> [Term]
>> id: MS:1000559
>> name: spectrum type
>> def: "Spectrum type." [PSI:MS]
>> relationship: part_of MS:1000442 ! spectrum
>>     
>
> All the children of the old "spectrum type" are really mass spectrum 
> types, so they would be children of the renamed "mass spectrum type":
>   
>> name: MS1 spectrum
>> name: MSn spectrum
>> name: CRM spectrum
>> name: SIM spectrum
>> name: SRM spectrum
>>     
>
> We don't have a term for PDA spectrum, UV spectrum, or ADC spectrum; I 
> suggest we add all these and make them children of the generic "spectrum 
> type" because unless I'm missing something they don't require their own 
> subtypes. Mass spectra definitely need their own subtype though (to ease 
> semantic validation: any mass spectrum should have m/z and intensity 
> arrays).
>
> Add a "chromatogram type" term.
> Redefine the "total ion chromatogram" term to be PART_OF "chromatogram 
> type".
> Add a "chromatogram" term analogous to the "spectrum" term (i.e. PART_OF 
> MS:0000000) and make "chromatogram type" PART_OF "chromatogram" just 
> like "spectrum_type" is PART_OF "spectrum".
> Add a "selected ion chromatogram" term.
> The "m/z" term could suffice to indicate which ion was "selected". But 
> for SRM SICs, an additional m/z is indicated to indicate the 
> product/fragment m/z and having two terms that just say "m/z" is 
> undesirable. Consequently, we might either add only a "product m/z" term 
> or add both "precursor m/z" and "product m/z" terms.
>
> That's all I can think of for now. :)
>
> -Matt
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
> Don't miss this year's exciting event. There's still time to save $100. 
> Use priority code J8TL2D2. 
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> _______________________________________________
> Psidev-ms-dev mailing list
> Psi...@li...
> https://lists.sourceforge.net/lists/listinfo/psidev-ms-dev
>