Hello,
I have some questions concerning protein groups derived from
ProteinProphet. The questions seem to be independent of recent TPP
versions (they arise as well with TPP v4.1 as with TPP v4.2). OS is
linux.
To illustrate my questions I have uploaded three XML extracts from two
different experiments with yeast & Sigma49 protXML files.
(1) How is a <protein_group> defined?
I figured out that all proteins in such a group are somehow connected
through shared peptides, but I don't understand how these protein
groups are defined.
My question is when are proteins with shared peptides put in the same
group and when are they put in separate ones?
There are 2 examples to illustrate my question in the uploaded
documents:
=> group703_yeast_ProteinProphet.xml holds three proteins belonging
to the same <protein_group>
These three proteins have only one peptide in common, namely
MLDMGFEPQIR.
=> group25.107_yeast_ProteinProphet.xml holds proteins with shared
peptides belonging to different
<protein_group> tags in the XML file.
In this case, the two different protein groups (containing each only
one protein) have 10 peptides in common and differ from each other
only in one peptide TGTKPAAVFAETLKHD for YDR012W in group 25 or two
peptides TGTKPAAVFTETLK and TGTKPAAVFTETLKHD for protein YBR031W in
group 107.
(2) How should the computed probabilities be interpreted?
What is the link between the probabilities computed for the groups and
the ones for the proteins?
Here again 2 examples:
=> group probability="1.0000", group holds 5 proteins, all proteins
have probability="0.0000". Does this mean we are sure that we saw one
of the five proteins, but we have absolutely no clue which one?
=> group probability="1.0000", group holds 5 proteins. First protein
has probability="0.9975", second one has probability="0.9955", the
other proteins have probability="0.0000". Does this mean I am 100%
sure I saw protein1? Or does it mean I am 99.75% sure I saw protein1
and 99.55% sure I saw protein2?
(3) How can a group have a probability="1.000", when:
1. all the peptides related to the group have
is_contributing_evidence="N"?
2. all proteins in the group have total_number_peptides="0"? and
3. all proteins in the group have probability="0"?
(see group38_sigma49_ProteinProphet.xml for such an example)
(4) If a peptide has is_contributing_evidence="N" for all its matching
proteins, what about the "minimal set" rule (want to find the smallest
set of proteins explaining all peptides)? If some peptides don't
contribute at all, how do you ensure that they will still be
"explained" by the final set of proteins?
I'm looking forward to your answers!
Sarah
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"spctools-discuss" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/spctools-discuss?hl=en
-~----------~----~----~----~------~----~------~--~---
