Re: [Psidev-ms-dev] amino acid modifications in TraML

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

> From: Andreas Bertsch [mailto:be...@in...]
>
> Dear mzTargetML community ;),
>
> during the implementation of the current draft version, some questions
> and comments came up.
>
> Cheers,
> A.
>
> 1. Currently, peptide modifications in traML are specified using the
> mass only. This can lead to ambiguous meanings.
>
>   <peptide id="ADTHFLLNIYDQLR-M1" proteinRef="Q12149">
>        <cvParam cvRef="MS" accession="MS:1000888" name="unmodified
> peptide sequence" value="ADTHFLLNIYDQLR"/>
>        <cvParam cvRef="MS" accession="MS:1000889" name="modified
> peptide sequence" value="ADTHFLLNIYDQLR[162.10111]"/>
>
> The same problem has been solved by the PI group in mzIdentML using
> the modification definitions of UniMod (snippet from the
> Macot_MSMS_example.mzid):
> .....
>   <Peptide id="peptide_3_9">
>        <peptideSequence>MSKPAGSTSRILDIPCK</peptideSequence>
>        <Modification location="0" monoisotopicMassDelta="127.063324">
>          <cvParam accession="UNIMOD:29" name="SMA" cvRef="UNIMOD" />
>        </Modification>
>        <Modification location="1" residues="M"
> monoisotopicMassDelta="15.994919">
>          <cvParam accession="UNIMOD:35" name="Oxidation"
> cvRef="UNIMOD" />
>          .....
>        </Modification>
>      </Peptide>
> .....
>
> If the modification itself is not known, the following term can be
> used (and is allowed in mzIdentML)
>
> [Term]
> id: MS:1001460
> name: unknown modification

A lot more cumbersome. But yes, I think this is probably better. I would
agree with this change.

> 2.  Modeling the peptide sequence using a CV term seems odd to me. I
> think this can be an attribute, or like in mzIdentML modeled as a
> subtag of peptide. The sequence must be given, I suppose, otherwise it
> can be modeled as a (unkown) compound.

Neutral on this. One reason for the current way of doing things was to
differentiate between modified and unmodified. With the above proposed
mechanism for modifications, this becomes less useful. The second reason
was mainly a general switch to everything as a CV term.

I'm happy to return to an attribute.

Other thoughts?

> 3 . Additionally, the proteinRef is not necessarily restricted to one
> single protein. There might be several proteins that contain that
> peptide and it should be possible to represent that. Make that a CV
> term?

This topic has surfaced before but never really dealt with well.

One reason to keep it an attribute is to insure proper reference to the
<protein> tag. That can't be done with a CV param (easily?).

One alternate possibility is to do something like:

<peptide id="ADTHFLLNIYDQLR-M1" sequence="ADTHFLLNIYDQLR">
  <proteinRef ref="Q12149">
  <proteinRef ref="P03934">
  <proteinRef ref="IPI00343833">
  <cvParam.....
</peptide>

With modification information:

<peptide id="ADTHFLLNIYDQLR-M1" sequence="ADTHFLLNIYDQLR">
  <proteinRef ref="Q12149">
  <proteinRef ref="P03934">
  <proteinRef ref="IPI00343833">
  <modification location="0" monoisotopicMassDelta="127.063324">
    <cvParam accession="UNIMOD:29" name="SMA" cvRef="UNIMOD"/>
  </modification>
  <modification location="1" residues="M"
monoisotopicMassDelta="15.994919">
    <cvParam accession="UNIMOD:35" name="Oxidation" cvRef="UNIMOD"/>
  </modification>
</peptide>

What do y'uns think of that?

Thanks,
Eric



>
>
> --
> 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
>
>
>
>
>
> -----------------------------------------------------------------------
> -------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart
> your
> developing skills, take BlackBerry mobile applications to market and
> stay
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Psidev-ms-dev mailing list
> Psi...@li...
> https://lists.sourceforge.net/lists/listinfo/psidev-ms-dev