search

mwalzer / psi-pi

Automatically exported from code.google.com/p/psi-pi
0 stars 0 forks source link

Decide how to specify modification locations for PMF #3

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
Consider a peptide with two methionine residues at residues 3 and 5 and a
mass such as only one should be modified.  How do we specify the location?
Could just be left blank, or could try and use a regular expression. Favour
the former, consider:
  ACBCCCDCECFCGCK
with 2 cys residues modified. The regular expression would be quite complex

Original issue reported on code.google.com by dcre...@gmail.com on 28 Apr 2008 at 12:44

GoogleCodeExporter commented 9 years ago 
In the current schema (June 12th) is a CustomModification element (for 
identified
peptides):

<CustomModification residue="" avgMassDelta="" location="" 
monoisotopicMassDelta="">
  <pf:cvParam accession="" name="" value="" cvRef=""/>
</CustomModification>

By the way: "location" and "residue" are both integer; I think "residue" should 
be
string. 

If we add an optional "count" attribute, would that model the identified PMF 
peptide?:

<Peptide identifier="peptide_3_1" sequenceMass="1945.920349" 
sequenceLength="15">
  <CustomModification count="2" residue="M" monoisotopicMassDelta="16.0">
    <pf:cvParam accession="" name="" value="" cvRef=""/>
  </CustomModification>
  <peptideSequence>NQVAMNPMNTIMDAK</peptideSequence>
</Peptide>

Original comment by eisena...@googlemail.com on 19 Jun 2008 at 12:14

GoogleCodeExporter commented 9 years ago 
Agreed in TeleCon on July 10th, 2008 to change it like proposed above.

Original comment by eisena...@googlemail.com on 10 Jul 2008 at 4:23

GoogleCodeExporter commented 9 years ago 
BUT: If same modification on different residues, how to handle the count?:
oxidation found 3 times in total, at M or S, how to specify?

wrong is (because we don't know the counts for the residues):
<Peptide id="pep1">
<CustomModification residue="M" count="1">
    <pf:cvParam accession="" name="oxidation" cvRef=""></pf:cvParam>
</CustomModification>
<CustomModification residue="S" count="2">
    <pf:cvParam accession="" name="oxidation" cvRef=""></pf:cvParam>
</CustomModification>
<peptideSequence>YHTSQSGDEMTSLSMYVSR</peptideSequence>
/Peptide>

Original comment by eisena...@googlemail.com on 10 Jul 2008 at 4:25

GoogleCodeExporter commented 9 years ago 
suggestion from TeleCon July 17th, 2008:
Drop the residue attribute and leave it to the CV to specify the locations
(especially PSI-MOD is quite exhaustive).

Then a correct solution for the above problem would be:
<Peptide id="pep1">
<CustomModification count="3">
    <pf:cvParam accession="PSI-MOD:xxxxx" name="oxidation" cvRef="PSI-MOD"/>
</CustomModification>
<peptideSequence>YHTSQSGDEMTSLSMYVSR</peptideSequence>
</Peptide>
with the CV term PSI-MOD:xxxxx containing the correct residues.

Original comment by eisena...@googlemail.com on 17 Jul 2008 at 3:58

GoogleCodeExporter commented 9 years ago 
I kind of like the solution of referring to PSI-MOD. But none of the examples 
use the
location attribute, which seems like a big loss of metadata. Was that just by
accident? I was also going to ask if the count attribute refers to the number of
times that the modification was applied to the specified location. That seems
appropriate for oxidation to methionine which can happen multiple times on the 
same
residue (or so I've been told).

Original comment by matthew....@vanderbilt.edu on 26 Aug 2008 at 10:17

GoogleCodeExporter commented 9 years ago 
The MS/MS examples use the location attribute.
The count attribute was not originally intended for MS/MS, but it can be used to
model more than one modification on one amino acid (quite rare):

<Modification location="1" residues="M">
 [PSI-MOD for Oxidation or Oxidation(M)]
</Modification>

The count attribute was intended for PMF, where we have no location, but only a 
possible residues:
<Modification residues="M" count="3">
 [PSI-MOD for Oxidation or Oxidation(M)]
</Modification>

or more than one:
<Modification residues="M" count="3">
 [PSI-MOD for Oxidation or Oxidation(M)]
</Modification>
<Modification residues="C" count="2">
 [PSI-MOD for Carbamido.]
</Modification>

Unfortunately in the current schema version the count is gone. We should add it 
again!

Original comment by eisena...@googlemail.com on 9 Oct 2008 at 2:05

GoogleCodeExporter commented 9 years ago 
TeleCon 9th of October:

leave count attribute dropped, use more instances of <Modification>
if there are more than one modifications.

Original comment by eisena...@googlemail.com on 9 Oct 2008 at 4:10

GoogleCodeExporter commented 9 years ago 
final example:

two oxidation modifications on residues M or W, one carbamidomethylation on C:

<Peptide id="bla">
 <Modification residues="MW">
  <pf:cvParam accession="MOD:00412" name="oxidation" cvRef="PSI-MOD"/>
 </Modification>
 <Modification residues="MW">
  <pf:cvParam accession="MOD:00412" name="oxidation" cvRef="PSI-MOD"/>
 </Modification>
 <Modification residues="C">
  <pf:cvParam accession="MOD:00397" name="iodoacetamide derivative" cvRef="PSI-MOD"/>
 </Modification>
 <peptideSequence>MIVWKKECNMCWTARPDEK</peptideSequence>
</Peptide>

David will create a PMF example, then this issue can be changed to Fixed.

Original comment by eisena...@googlemail.com on 9 Oct 2008 at 5:03

GoogleCodeExporter commented 9 years ago

Original comment by eisena...@googlemail.com on 23 Oct 2008 at 4:01