mono or average masses

[GoogleCodeExporter]

All mass values, e.g. sequenceMass of <Peptide> or calculatedMassToCharge,
experimentalMassToCharge of <SpectrumIdentificationItem> are not specified
whether monoisotopic or averaged.

We decided to have only one type of masses in the whole analysisXML.
But there is no "global" schema attribute.

One can specify a CVparam for a <SpectrumIdentificationProtocol> (e.g.
<pf:cvParam accession="PRIDE:0000162" name="Mass value type setting
monoisotopic" cvRef="PRIDE"/>), but then it could differ between
<SpectrumIdentificationProtocol>s in the same analysisXML).

Original issue reported on code.google.com by eisena...@googlemail.com on 30 Jul 2008 at 12:13

[GoogleCodeExporter]

From Martin:

> I think the current schema is not okay, because it allows "average" in one
SpecIdent and "mono" in another,
> so it is not well-defined for the masses in elements or attributes.
> We need a global attribute :-) or element. Or it can be done later in semantic
validation :-( .

Original comment by dcre...@gmail.com on 7 Aug 2008 at 1:55

[GoogleCodeExporter]

I think it's actually _required_ to be like this. For example, at least one 
search
engine allows you to specify mono for masses below x and average for masses 
above x.
So, in this case, the output should be similar to the N15 example that I've 
supplied,
with two separate mass tables. Maybe you could look at the 
Mascot_N15_example.xml and
see if you think that this is OK. 

Original comment by dcre...@gmail.com on 7 Aug 2008 at 1:56

[GoogleCodeExporter]

Um... what about average peptide masses and monoisotopic fragment masses - or 
vice versa.

Original comment by dcre...@gmail.com on 7 Aug 2008 at 1:59

[GoogleCodeExporter]

regarding comment3: that could be another type of search, e.g. "mass type 
setting
PepAvg FragMono" as child CV term of "mass type settings" next to "mass type 
setting
average isotopic" and "mass type setting monoisotopic".

Oh, I get the point. There is no mechanism to reference two mass tables in one
search...!?

But is that type of search common? We should close this issue for milestone 1...

Original comment by eisena...@googlemail.com on 7 Aug 2008 at 2:17

[GoogleCodeExporter]

Summary: 
Different SpectrumIdentification runs may have different (e.g. average and 
mono) mass
tables.

Remaining problems:
1) giving the mass of a peptide may then duplicate the sequence information 
(when a
peptide is identified in the avg AND in the mono run)
possible solution: drop any mass information and give it only in relation to a 
run
(i.e. in the PeptideEvidence or SpectrumIdentificationItem elements)
2) if parent masses are avg and fragment masses are mono (and fragments are
reported); at the moment we have only one mass table!
3) if masses below X Da are avg and above X Da are mono

Original comment by eisena...@googlemail.com on 1 Sep 2008 at 12:57

[GoogleCodeExporter]

The mono / average problem is solved meanwhile:
because we have a MassTable_ref from the <Peptide> elements and the
possibility to have more than one mass table per 
<SpectrumIdentificationProtocol>.

A peptide with the same sequence, but another mass table has to
be reported twice (or more often), but that is a bitter pill we have 
to swallow

Original comment by eisena...@googlemail.com on 9 Oct 2008 at 11:44

• Changed state: Fixed