[Psidev-pi-dev] N15 example

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

Hi David,

I'm trying to understand how best to model differential labelling e.g. the N15 file. I think we should view this as similar to searching for different variable mods. For example, if this was an ICAT, we would search for C+ICATlight and C+ICAT heavy within one single search. As such, I don't think it makes sense to separate out an N15 labelling into two different search protocols.

I think we should allow multiple mass tables within one SpectrumIdentificationProtocol. The MassTable also will need an id and an optional cv terms to explain the purpose of this particular mass table and perhaps to say if it is monoisotopic or average. We could then add a MassTable_ref to Peptide, documented saying that it is only required if more than one MassTable has been given.

If we also have:

<FragmentationTable>
          <Measure id="m_mz">
            <pf:cvParam cvRef="PSI-PI" accession="PI:xxxx" name="product ion monoisotopic m/z"/>

To say that reported fragment ions are monoisotopic, does this also solve the problem if someone wants to report that precursor ions are average mass and product ions are monoisotopic?


Just to be clear however, in SpectrumIdentificationProtocol we have:

<SpectrumIdentificationProtocol id="SIP_1" AnalysisSoftware_ref="AS_mascot_server">
      <AdditionalSearchParams>
	
        ...
        <pf:cvParam accession="PI:00211" name="mass type setting monoisotopic"     cvRef="PSI-PI"/>

I didn't understand why we couldn't have two CV terms here for product and precursor mass setting, this is how X!Tandem sets up its searches. Was there something else that I missed here that would stop this working?
Cheers
Andy