Re: [Psidev-ms-dev] precursor/product/neutral loss/SRM/MSE/MS3 scans

[Matt C. <mat...@va...>]

Hi Fredrik,

The new examples for precursor ion scans and neutral loss scans look 
good. I'm not sure about the names "product" and "neutralLoss" because 
they seem to refer to a single ion as opposed to the configuration of 
the scan. But I don't have any better ideas at the moment. Also, can we 
reasonably expect future fantastic configurations (like a forked triple 
quad which splits after Q2 to two Q3s, for example) to scan for multiple 
neutral losses from a given initial ion? And is it possible to compose a 
precursor ion scan intensity from multiple Q3 products?

As far as the keyrefs go, I put them because ID and IDREF merely 
validate that any id is pointed to, not an id of the appropriate element 
type. For example, with just ID/IDREF, a spectrumRef could refer to a 
sourceFile's id attribute. That's obviously undesirable and with 
key/keyref it's easily avoidable and validates with just the schema. 
Thanks to Angel for showing me how to use them.

-Matt


Eric Deutsch wrote:
> Hi Fredrick, thank you for the summary. This looks very good to me. Let's
> discuss and approve at tomorrow's call if we can get a quorum.
>
> Regarding the keyrefs, you just removed some keyrefs that were already of
> type ID or IDREF and therefore don't need the keyref?
>
> Thanks!
> Eric
>
>
>   
>> -----Original Message-----
>> From: Fredrik Levander [mailto:Fre...@im...]
>> Sent: Tuesday, December 23, 2008 12:59 AM
>> To: Mass spectrometry standard development
>> Subject: Re: [Psidev-ms-dev] precursor/product/neutral loss/SRM/MSE/MS3
>> scans
>>
>> I think holidays and snow are hampering feedback, so far, but I could
>> summarize the proposed schema changes to make it easier to follow.
>> Note that it will not change anything for product scans or normal MS
>> scans (apart from the xmlns and version), the tiny example validates
>> without problem.
>>
>> Summary of changes:
>>
>>  - Target namespace changed to allow schema to reside there. Version
>> updated to 1.1.0
>>  - scanList minOccurs="0"
>>  - in spectrum, allow either precursors, product or neutraLoss element:
>>           <xs:choice minOccurs="0" maxOccurs="1">
>>               <xs:element name="precursorList"
>> type="dx:PrecursorListType" />
>>               <xs:element name="product" type="dx:ProductType" />
>>               <xs:element name="neutralLoss" type="dx:NeutralLossType" />
>>           </xs:choice>
>>  - IsolationWindowType description generalized to accomodate all types
>> of isolated ions.
>>  - SelectedIonList description generalized to accomodate all types of
>> isolated ions
>>  - Addition of ProductType and NeutralLossType (containing activation
>> and isolationWindow, no selectIonList ).
>>  - Removed some keyrefs
>>
>> If this is OK the schema is found at:
>>
>> http://dev.thep.lu.se/fp6-
>> prodac/browser/trunk/mzML/scans/mzML_dev_scans.xsd
>>
>> and sample spectrum examples:
>> http://dev.thep.lu.se/fp6-prodac/browser/trunk/mzML/scans
>> (The link was not correct in the previous post)
>>
>> Thanks,
>>
>> Fredrik
>>
>> Fredrik Levander wrote:
>>     
>>> Since it is quite unlikely with a sequence of isolation events within
>>>       
>> one scan, I revised the scan examples, xsd and examples to only allow
>> either a precursorList, product or neutralLoss. This way there would not
>> be any changes to normal MS2 files.
>>     
>>> The updated files are found at the same URL.
>>> http://dev.thep.lu.se/fp6-prodac/browser/trunk/mzML/scans
>>>
>>> These examples are really minimalistic, and probably they are not
>>>       
>> semantically complete, but at least they validate with the xsd.
>>     
>>> thoughts?
>>>
>>> Fredrik
>>>
>>> ----- Original Message -----
>>> From: Fredrik Levander <Fre...@im...>
>>> Date: Thursday, December 18, 2008 5:02 pm
>>> Subject: [Psidev-ms-dev] precursor/product/neutral loss/SRM/MSE/MS3
>>>       
>> scans
>>     
>>>       
>>>> Hi All,
>>>>
>>>> To find out whether mzML is OK for various scan types, I've tried
>>>> to
>>>> write down some example scans of different types in mzML format. As
>>>> discussed previously, precursor and neutral loss scans didn't fit
>>>> well
>>>> into the schema. I've tried to make a minimal set of changes to the
>>>> schema to support different scan types.
>>>> I've uploaded scan examples (product / precursor / neutral loss /
>>>> MRM /
>>>> MSe / MS3) and an updated schema and sample file for discussion at:
>>>> http://dev.thep.lu.se/fp6-prodac/browser/trunk/mzML/scans
>>>>
>>>> In schema terms my proposal would be to generalise precursorList by
>>>> renaming it to isolationList and then allow elements of types
>>>> 'precursor', 'product' and 'neutralLoss', that are all of the
>>>> 'IsolationType', which is about the same as the previous
>>>> 'PrecursorType'. This would make it possible to accomodate quite
>>>> complex
>>>> combinations of scans.
>>>>
>>>> Other  observations:
>>>> - The isolation window width might not be known, should the center
>>>> mass
>>>> be given instead then? See the neutral loss example.
>>>> - scanList should probably be optional (as was acquisitionList and
>>>> scan
>>>> preciously)
>>>> - keyrefs are only necessary for spectrumId, other keys are checked
>>>> automatically since they are ID-IDREF.
>>>>
>>>> If anyone would like to download the files using subversion:
>>>> svn checkout  http://dev.thep.lu.se/fp6-prodac/svn/trunk/mzML/scans
>>>> scans
>>>> Thanks,
>>>>
>>>> Fredrik
>>>>
>>>>         
>