psidev-ms-dev

[Psidev-ms-dev] Is a <transition> a <target> with an additional <product> ?

[Steffen N. <sne...@ip...>]

Hi,

this might be a bit provocative, but why are <transition> 
and <target> separated ? 

Why not rename <transition> to <target>,
and make the <product> optional ? This comes from 
somebody who thinks of an MRM experiment 
as a "degraded" tandemMS measurement (no offense meant).

If you run tandemMS on peptides, you might still 
want to annotate the <TargetIncludeList> with 
peptides etc, just as now the <transition>s.

To get rid of the multiple toplevel <*List>
one could have a <TargetList role="include|exclude">,
and the <TargetList> could have maxOccurs=2,
but that could make life more difficult 
for the semantic validators.

Yours,
Steffen

-- 
IPB Halle                    AG Massenspektrometrie & Bioinformatik
Dr. Steffen Neumann          http://www.IPB-Halle.DE
Weinberg 3                   http://msbi.bic-gh.de
06120 Halle                  Tel. +49 (0) 345 5582 - 1470
                                  +49 (0) 345 5582 - 0
sneumann(at)IPB-Halle.DE     Fax. +49 (0) 345 5582 - 1409

Re: [Psidev-ms-dev] Is a <transition> a <target> with an additional <product> ?

[Steffen N. <sne...@ip...>]

Hi,

maybe this suggestion drowned in the mail traffic 
that had accumulated over the holidays, 
so I'll bring this up again. 

The suggestion is to merge <transition> and <target>
into a single concept <target>, where the <product> is optional.

Comments ?

Yours,
Steffen

On Tue, 2010-01-05 at 23:09 +0100, Steffen Neumann wrote:
> Hi,
> 
> this might be a bit provocative, but why are <transition> 
> and <target> separated ? 
> 
> Why not rename <transition> to <target>,
> and make the <product> optional ? This comes from 
> somebody who thinks of an MRM experiment 
> as a "degraded" tandemMS measurement (no offense meant).
> 
> If you run tandemMS on peptides, you might still 
> want to annotate the <TargetIncludeList> with 
> peptides etc, just as now the <transition>s.
> 
> To get rid of the multiple toplevel <*List>
> one could have a <TargetList role="include|exclude">,
> and the <TargetList> could have maxOccurs=2,
> but that could make life more difficult 
> for the semantic validators.
> 
> Yours,
> Steffen
> 


-- 
IPB Halle                    AG Massenspektrometrie & Bioinformatik
Dr. Steffen Neumann          http://www.IPB-Halle.DE
Weinberg 3                   http://msbi.bic-gh.de
06120 Halle                  Tel. +49 (0) 345 5582 - 1470
                                  +49 (0) 345 5582 - 0
sneumann(at)IPB-Halle.DE     Fax. +49 (0) 345 5582 - 1409

Re: [Psidev-ms-dev] Is a <transition> a <target> with an additional <product> ?

[Steffen N. <sne...@ip...>]

On Wed, 2010-01-13 at 13:12 +0100, Andreas Bertsch wrote:
> I agree with Fredrik. The costs for the simplifications are too high.
Yes, it's not for free :-(

> Separating the targets and transitions seems reasonable to me, because
> these are to very different things.
Are they ? 

Analytically yes, because you program the MS machine 
in different ways, either (auto-)MS/MS or SRM.

Conceptionally no, because if I have observed a pair 
of precursor/product ions and use that as a hint that 
the predicted interpretation is true, no matter whether 
you have measured that by SRM or a normal QQQ 
or QqTOF tandemMS experiment. Granted, only the <precursor> 
information will not be enough for a usable <interpretation>.

On the other hand, to analyse a normal tandemMS experiment,
you could even specify multiple <product>s for one precursor,
which you require for a sufficiently confident peptideRef="ABC".
(If I saw that correctly from the traML schema, the current approach
 is to create multiple <transition> to hold multiple <product>s)

Again, I think I can be persuaded/convinced if there 
is a definition of use case(s) for traML, and my ideas 
fall well outside of it. Has anyone start that already ?

> The subelements are reused for both of the lists, which is I think a
> good trade-off between a light-weight schema and straight-forward
> structure.
And that even allows the mapping file to distinguish between 
a retention time inside a <transition> vs. a retention time 
in a <target>. But that should not be the main point.

Yours,
Steffen

-- 
IPB Halle                    AG Massenspektrometrie & Bioinformatik
Dr. Steffen Neumann          http://www.IPB-Halle.DE
Weinberg 3                   http://msbi.bic-gh.de
06120 Halle                  Tel. +49 (0) 345 5582 - 1470
                                  +49 (0) 345 5582 - 0
sneumann(at)IPB-Halle.DE     Fax. +49 (0) 345 5582 - 1409

Re: [Psidev-ms-dev] Is a <transition> a <target> with an additional <product> ?

[Eric D. <ede...@sy...>]

Hi Steffen, Andreas, Fredrik, et al. thank you for this discussion on
Steffen's proposal. The history here is that we set out to create a format
for transitions only. The primary use cases all were for SRM experiments.
This has not yet been clearly written in the spec doc as you noted. But
then it was suggested that we could easily add inclusion/exclusion lists
as well. It seemed unclear to me that there would be much interest in
sharing inclusion lists (in my mind, transition lists have great reuse
potential, but inclusion lists far less so), but it seemed reasonable to
add this on to permit certain additional use cases as there was no
alternative and this would be the closest similar format. However, I, too,
am hesitant to complicate the <transition> design (and hinder validation,
e.g. how do you prevent transitions with no Q3 value?) by trying to fit
this additional potential use case into our current design.

Steffen, thank you for writing the use case section. I will insert this
into the spec doc and perhaps expand on it a bit.

If there are addition comments, please do speak up. Let's also entertain
this idea bit more on Tuesday's phone conference.

Thanks,
Eric


> -----Original Message-----
> From: Steffen Neumann [mailto:sne...@ip...]
> Sent: Wednesday, January 13, 2010 5:52 AM
> To: Mass spectrometry standard development
> Subject: Re: [Psidev-ms-dev] Is a <transition> a <target> with an
> additional <product> ?
>
> On Wed, 2010-01-13 at 13:12 +0100, Andreas Bertsch wrote:
> > I agree with Fredrik. The costs for the simplifications are too high.
> Yes, it's not for free :-(
>
> > Separating the targets and transitions seems reasonable to me,
> because
> > these are to very different things.
> Are they ?
>
> Analytically yes, because you program the MS machine
> in different ways, either (auto-)MS/MS or SRM.
>
> Conceptionally no, because if I have observed a pair
> of precursor/product ions and use that as a hint that
> the predicted interpretation is true, no matter whether
> you have measured that by SRM or a normal QQQ
> or QqTOF tandemMS experiment. Granted, only the <precursor>
> information will not be enough for a usable <interpretation>.
>
> On the other hand, to analyse a normal tandemMS experiment,
> you could even specify multiple <product>s for one precursor,
> which you require for a sufficiently confident peptideRef="ABC".
> (If I saw that correctly from the traML schema, the current approach
>  is to create multiple <transition> to hold multiple <product>s)
>
> Again, I think I can be persuaded/convinced if there
> is a definition of use case(s) for traML, and my ideas
> fall well outside of it. Has anyone start that already ?
>
> > The subelements are reused for both of the lists, which is I think a
> > good trade-off between a light-weight schema and straight-forward
> > structure.
> And that even allows the mapping file to distinguish between
> a retention time inside a <transition> vs. a retention time
> in a <target>. But that should not be the main point.
>
> Yours,
> Steffen
>
> --
> IPB Halle                    AG Massenspektrometrie & Bioinformatik
> Dr. Steffen Neumann          http://www.IPB-Halle.DE
> Weinberg 3                   http://msbi.bic-gh.de
> 06120 Halle                  Tel. +49 (0) 345 5582 - 1470
>                                   +49 (0) 345 5582 - 0
> sneumann(at)IPB-Halle.DE     Fax. +49 (0) 345 5582 - 1409
>
>
> -----------------------------------------------------------------------
> -------
> This SF.Net email is sponsored by the Verizon Developer Community
> Take advantage of Verizon's best-in-class app development support
> A streamlined, 14 day to market process makes app distribution fast and
> easy
> Join now and get one step closer to millions of Verizon customers
> http://p.sf.net/sfu/verizon-dev2dev
> _______________________________________________
> Psidev-ms-dev mailing list
> Psi...@li...
> https://lists.sourceforge.net/lists/listinfo/psidev-ms-dev

Re: [Psidev-ms-dev] Is a <transition> a <target> with an additional <product> ?

[Fredrik L. <Fre...@im...>]

Hi Steffen,

I think your suggestion is fine and should be working. 
InterpretationList/Interpretation, now part of  Transition, would be 
explaining different things for included ions and transitions, though.

Even if such a change would simplify the XML schema which is good, I am 
afraid that it would make reading of the documents a bit harder - you 
would still like to distinguish between include targets and transitions, 
I believe.  So in the end the current representation with separate 
transition and target elements seems preferable to me. Or are there any 
other practical advantages with merging of the two? Note that targets in 
the include list can already be annotated with peptide information - 
actually they have to be, in order to have a retention time.

Thanks

Fredrik


Steffen Neumann wrote:
> Hi,
>
> maybe this suggestion drowned in the mail traffic 
> that had accumulated over the holidays, 
> so I'll bring this up again. 
>
> The suggestion is to merge <transition> and <target>
> into a single concept <target>, where the <product> is optional.
>
> Comments ?
>
> Yours,
> Steffen
>
> On Tue, 2010-01-05 at 23:09 +0100, Steffen Neumann wrote:
>   
>> Hi,
>>
>> this might be a bit provocative, but why are <transition> 
>> and <target> separated ? 
>>
>> Why not rename <transition> to <target>,
>> and make the <product> optional ? This comes from 
>> somebody who thinks of an MRM experiment 
>> as a "degraded" tandemMS measurement (no offense meant).
>>
>> If you run tandemMS on peptides, you might still 
>> want to annotate the <TargetIncludeList> with 
>> peptides etc, just as now the <transition>s.
>>
>> To get rid of the multiple toplevel <*List>
>> one could have a <TargetList role="include|exclude">,
>> and the <TargetList> could have maxOccurs=2,
>> but that could make life more difficult 
>> for the semantic validators.
>>
>> Yours,
>> Steffen
>>
>>     
>
>
>   

Re: [Psidev-ms-dev] Is a <transition> a <target> with an additional <product> ?

[Steffen N. <sne...@ip...>]

On Wed, 2010-01-13 at 11:47 +0100, Fredrik Levander wrote:
> InterpretationList/Interpretation, now part of  Transition, would be 
> explaining different things for included ions and transitions, though.
The <nterpretation> would move into <product> then, 
because really it's describing the <precursor>:<product> interpretation.

> Even if such a change would simplify the XML schema which is good, I am 
> afraid that it would make reading of the documents a bit harder - you 
> would still like to distinguish between include targets and transitions.
What kind of format is TraML supposed to be ? The "Use Case" section 
of the documentation is pretty sparse so far. Where in the pipeline 
is TraML ? I'd need to know to avoid asking for something in traML 
it was never intended to do. 

Do you have a list of possible proteins/peptides you want to 
look for / quantitate ? And hence create a <transition> 
to be imported/converted into the machine software 
to run the SRM experiment ? So traML describes the settings 
for RT,Q1,Q3.  

With separate <transition> and <target>
there is a strong hint that it is designed for SRM capable 
machines only, i.e. *not* for QTOF. 

> Or are there any other practical advantages with merging of the two? 
With a unified <target> traML you can describe RT,Q1 and optionally Q3.
It'd make a much more universal description of MS experiments. 

> Note that targets in 
> the include list can already be annotated with peptide information - 
> actually they have to be, in order to have a retention time.
Which is another thing that struck me, why do I need 
the peptide information to store a retention time ? 

Depending on the Use Case I don't have peptide/compound information,
it might be that I am actually making the MSMS experiment 
to obtain that!

Yours,
Steffen



-- 
IPB Halle                    AG Massenspektrometrie & Bioinformatik
Dr. Steffen Neumann          http://www.IPB-Halle.DE
Weinberg 3                   http://msbi.bic-gh.de
06120 Halle                  Tel. +49 (0) 345 5582 - 1470
                                  +49 (0) 345 5582 - 0
sneumann(at)IPB-Halle.DE     Fax. +49 (0) 345 5582 - 1409

Re: [Psidev-ms-dev] Is a <transition> a <target> with an additional <product> ?

[Andreas B. <be...@in...>]

Hi Steffen,

I agree with Fredrik. The costs for the simplifications are too high. Separating the targets and transitions seems reasonable to me, because these are to very different things. The subelements are reused for both of the lists, which is I think a good trade-off between a light-weight schema and straight-forward structure.

Cheers,
A.

> I think your suggestion is fine and should be working. 
> InterpretationList/Interpretation, now part of  Transition, would be 
> explaining different things for included ions and transitions, though.
> 
> Even if such a change would simplify the XML schema which is good, I am 
> afraid that it would make reading of the documents a bit harder - you 
> would still like to distinguish between include targets and transitions, 
> I believe.  So in the end the current representation with separate 
> transition and target elements seems preferable to me. Or are there any 
> other practical advantages with merging of the two? Note that targets in 
> the include list can already be annotated with peptide information - 
> actually they have to be, in order to have a retention time.
>> 
>> maybe this suggestion drowned in the mail traffic 
>> that had accumulated over the holidays, 
>> so I'll bring this up again. 
>> 
>> The suggestion is to merge <transition> and <target>
>> into a single concept <target>, where the <product> is optional.
>> 
>> Comments ?
>>> this might be a bit provocative, but why are <transition> 
>>> and <target> separated ? 
>>> 
>>> Why not rename <transition> to <target>,
>>> and make the <product> optional ? This comes from 
>>> somebody who thinks of an MRM experiment 
>>> as a "degraded" tandemMS measurement (no offense meant).
>>> 
>>> If you run tandemMS on peptides, you might still 
>>> want to annotate the <TargetIncludeList> with 
>>> peptides etc, just as now the <transition>s.
>>> 
>>> To get rid of the multiple toplevel <*List>
>>> one could have a <TargetList role="include|exclude">,
>>> and the <TargetList> could have maxOccurs=2,
>>> but that could make life more difficult 
>>> for the semantic validators.

--
Div. for Simulation of Biological Systems, WSI, University of Tuebingen
Room C322, Sand 14, 72076 Tuebingen, Germany
phone: +49-7071-29-70461 fax: +49-7071-29-5152
http://www-bs.informatik.uni-tuebingen.de