Re: [Psidev-pi-dev] Modifications

[Jones, A. <And...@li...>]

I think the count attribute is just confusing, if two modifications have been identified, just have two entries of the Modification element? 

I don't personally find the count attribute confusing, but don't have a strong preference. Also, see next item.



There would still be an issue of how to specify which residue has been modified but basically this is the same issue for standard and custom mods i.e. the CV tells you the mod and the residue?

Not necessarily. Suppose that you have 2 phosphorylated residues, PMF and the peptide is STYSTYSTYK
I still don't see how to encode this?

 

These two things are sort of related in that I’d like to only have one mechanism for describing mods if possible. My suggestion is that it is done as follows:

 

<Modification>

                <cvParam name = “phosphorylation”/>

</Modification>

<Modification>

                <cvParam name = “phosphorylation”/>

</Modification>

 

Since from PMF this is all the information we have i.e. we don’t know the location or the residue (we may have a mass delta value though?).

 

If we know the residues modified, then it is encoded as follows:

 

<Modification location = “3”>

                <cvParam name = “phosphorylation to tyrosine”/>

</Modification>

<Modification location = “6”>

                <cvParam name = “phosphorylation to tyrosine”/>

</Modification>

 

 

I agree this use of CV terms doesn’t appear ideal but the current schema allows you to encode the same information in lots of different ways, say two tyrosines are phosphorylated:

 

<CustomMod count = “2” residue = “Y”>

                <cvParam name = “phosphorylation”/>

</CustomMod>

 

OR

 

<Modification>

                <cvParam name = “phosphorylation to tyrosine”/>

</Modification>

<Modification>

                <cvParam name = “phosphorylation to tyrosine”/>

</Modification>

 

OR

 

<CustomMod count = “2” >

                <cvParam name = “phosphorylation to tyrosine”/>

</CustomMod>

 

I think we ought to discuss the pros and cons of encoding the mod and the residue modified in one single CV term – it should work in theory but it seems like more work for parsers than just having an attribute for residue = “Y”...

 

Cheers, Andy

 

 

 

 

 

From: David Creasy [mailto:dc...@ma...] 
Sent: 30 September 2008 10:55
To: psi...@li...
Cc: psi...@li...
Subject: Re: [Psidev-pi-dev] Modifications

 

Hi,

Comments in line below:

Jones, Andy wrote: 

Hi all,

 

I’d like to get the mods part of the schema sorted in time for the call this week if possible. There are quite a few different aspects in issues 3 and 35 so I’ll try to summarise it here:

 

In the SpectrumIdentificationProtocol we have:

 

        <SearchModification fixedMod="false" >

          <ModName   accession="MOD:TODO" name="SMA (N-term)" cvRef="PSI-MOD" />

          <MassValue value="127.063324" unitAccession="PSI:xxxx" unitName="Da" />

          <SpecificityRule accession="" cvRef="" name="" unitAccession="" unitName="" />

        </SearchModification>

 

On Peptide we have:

 

      <Modification location="13">

        <pf:cvParam accession="TODO: (requires quite a bit of code)" name="SMA (K)" cvRef="PSI-MOD" />

      </Modification>

 

OR

 

      <SubstitutionModification originalResidue="K" replacementResidue="M" location="2">

        <pf:cvParam cvRef="" accession="" name=""/>

      </SubstitutionModification>

 

(Note the datatypes were set to int for original and replaceResidue so I have fixed this to be the same alphabet as for pre and post).

 

OR

 

      <CustomModification location="2" monoisotopicMassDelta="21.21" count="1" residue = “M”>

        <pf:cvParam cvRef="" accession="" name=""/>

      </CustomModification>

 

 

PSI-MOD can specify a modification and the residue that has been modified, although the terms are not always intuitive for our purposes. For example, as far  as I can tell “oxidation to L-methionine sulfoxide” would be the standard term for a methionine oxidation, correct?

Um, I don't think so. It looks to me as though Oxidation of Methionine is possibly missing PSI-MOD at the moment...
I would expect there to be a DiffFormula: "O 1" and from unimod the PSI-MS name is "Oxidation".
So, it looks as though something may be wrong here... I've cc'd psi...@li... so hopefully someone will correct/enlighten us.
(The .obo file hasn't been updated since 20th April, and I know that there have been changes since then.)




 

Comments on current schema:

I’m not sure we need CustomModification, I would prefer just to have the Modification element, making location optional and adding the mass delta attributes to Modification.

Sounds fine. And the CV would be optional too?



I think the count attribute is just confusing, if two modifications have been identified, just have two entries of the Modification element? 

I don't personally find the count attribute confusing, but don't have a strong preference. Also, see next item.



There would still be an issue of how to specify which residue has been modified but basically this is the same issue for standard and custom mods i.e. the CV tells you the mod and the residue?

Not necessarily. Suppose that you have 2 phosphorylated residues, PMF and the peptide is STYSTYSTYK
I still don't see how to encode this?



I disagree that we should use the same element for the searched and the found modification. We do not need to report the specificity on the found modifications and in most cases you would not report the mass delta for a found modification (presumably this would only make any sense for MS1 data?)

OK



On SearchModification, are we agreed that we want to report the MassValue (i.e. mass of residue +/- mod) rather than say a MassDelta (mass of mod only)?

No! I think a delta is better. Suppose that you are performing an 14N/15N experiment and want to use Oxidation. That would mean having two separate SearchModifications.
(Also, I don't think that PSI-MOD should have this, but that's a separate issue...)



I think we should add documentation that MassValue is optional since the mod mass is part of the CV, but if reported the MassValue “overrides” the mass value from the CV.

SubstitutionMod inherits a mandatory association to Param, this needs to be changed since in most cases a CV term would not be required.

 

Thoughts?

Cheers, Andy

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


________________________________



 
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
 


________________________________



 
_______________________________________________
Psidev-pi-dev mailing list
Psi...@li...
https://lists.sourceforge.net/lists/listinfo/psidev-pi-dev
  





-- 
David Creasy
Matrix Science
64 Baker Street
London W1U 7GB, UK
Tel: +44 (0)20 7486 1050
Fax: +44 (0)20 7224 1344
 
dc...@ma...
http://www.matrixscience.com
 
Matrix Science Ltd. is registered in England and Wales
Company number 3533898