Latest INOH Biopax file Re: couple more things about INOH Biopax
Ken Fukuda
It is very late or very early in the morning here.
So no documents for today. Sorry about this.
Here is the latest INOH BioPAX TGF-beta file.
Known issues:
*duplicated evidence instances
*duplicated pathwaystep instances
*publicated PPE instances.
Best,
Ken
2006/6/1, Alan Ruttenberg <alanrut...@gmail.com>:
>
> Hi Ken,
>
> Some more stuff:
>
> These are direct instances of owl:Thing, which usually means that
> they are referred to somewhere, but not defined anywhere (pellet says
> they must be *something* so it makes them instances of owl:Thing)
>
> Connected_id1657654683_id1285092580,
> Connected_id997589207_id46795497, Connected_id625026951_id1588707703,
> id699451362_Smad7, Connected_id121677340_id33496024, !<http://
> www.biopax.org/release/Proteasome%20degradation>,
> Connected_id1165069402_id1936035698, id1387703047_Transcription0,
> Connected_id1350648522_id114353630,
> Connected_id1165069402_id1709918560,
> Connected_id1350648522_id1588707703, id687446721_Transcription0,
> Connected_id997589207_id114353630,
> Connected_id1657654683_id1709918560, id559533476_Smad7,
> Connected_id625026951_id1200991115, Connected_id121677340_id933799551
>
> In addition there are many unification xrefs
>
> <unificationXref rdf:ID="IEV_0000010">
> <DB rdf:datatype="http://www.w3.org/2001/
> XMLSchema#string">IEV_0000010</DB>
> <ID rdf:datatype="http://www.w3.org/2001/
> XMLSchema#string">Proteasome degradation</ID>
> <COMMENT rdf:datatype="http://www.w3.org/2001/
> XMLSchema#string">http://www.inoh.org (ProcessOntology)</COMMENT>
> </unificationXref>
>
> In these, I think that the DB should be something like "INOH Process
> Ontology"(or better a definitive URL for the the process ontology),
> and the ID IEV_0000010 and the comment (or better the rdfs:label)
> "Proteasome degradation"
>
> Finally, if you add something like this:
>
> <!DOCTYPE rdf:RDF [
> <!ENTITY owl "http://www.w3.org/2002/07/owl#">
> <!ENTITY rdf "http://www.w3.org/1999/02/22-rdf-syntax-ns#">
> <!ENTITY rdfs "http://www.w3.org/2000/01/rdf-schema#">
> <!ENTITY xsd "http://www.w3.org/2001/XMLSchema#">
> ]>
>
> at the beginning of your file between the first two lines:
>
> <?xml version="1.0" encoding="UTF-8"?>
> ************** over here *************
> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>
> you will be able to abbreviate
>
> <DB rdf:datatype="http://www.w3.org/2001/
> XMLSchema#string">IEV_0000010</DB>
>
> to
>
> <DB rdf:datatype="&xsd;string">IEV_0000010</DB>
>
> to make the file a bit shorter.
>
> Later,
> Alan
>
>
> >
>
Alan Ruttenberg
Starting to look at the latest version.
Here's what I've found so far:
<bp:sequenceSite rdf:ID="id1657654683_id1709918560_ss">
<bp:POSITION-STATUS>CERTAIN</bp:POSITION-STATUS>
<bp:SEQUENCE-POSITION rdf:datatype="http://www.w3.org/2001/
XMLSchema#unsignedLong"/>
</bp:sequenceSite>
In this the value CERTAIN isn't one of the choices (EQUAL, LESS-THAN,
GREATER-THAN)
Also, the SEQUENCE-POSITION doesn't have a value (it needs one if
you are listing it. If the value is unknown you would simply not
state it). If there isn't a SEQUENCE-POSITION that is known, it's not
clear what the point of having the sequenceSite instance is.
As I noted in the previous version, PARTICIPANTS is still being
directly asserted.
In the following
<sequenceParticipant rdf:ID="id1657872740_MAPK0">
<ORGANISM rdf:resource="#Canonical"/>
<PHYSICAL-ENTITY rdf:resource="#MI0003570_MAPK"/>
<CELLULAR-LOCATION
rdf:resource="#plasma_membrane__extrinsic_to_internal_side_of_plasma_mem
brane_"/>
<EVIDENCE rdf:resource="#id1657872740_MAPK_evd1"/>
<XREF rdf:resource="#INOH_II0000028_id1657872740"/>
</sequenceParticipant>
ORGANISM is a property of rna, protein, pathway, complex, or dna, but
not participants.
EVIDENCE is a property on interactions or pathways, but not
participants
similar issue with XREF, which isn't valid on participants
A number of "dangling pointers" items are referenced, but not defined.
"Untyped Individual: Assuming biopax-
level2:Connected_id1657654683_id1709918560 is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id1350648522_id1588707703 is an individual"
"Untyped Individual: Assuming biopax-
level2:id1387703047_Transcription0 is an individual"
"Untyped Individual: Assuming http://www.biopax.org/release/biopax-
level2.owl# is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id1350648522_id114353630 is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id625026951_id1588707703 is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id625026951_id1200991115 is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id1165069402_id1936035698 is an individual"
"Untyped Individual: Assuming proteasome:20degradation is an individual"
"Untyped Individual: Assuming biopax-level2:id559533476_Smad7 is an
individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id997589207_id46795497 is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id1165069402_id1709918560 is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id121677340_id33496024 is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id997589207_id114353630 is an individual"
"Untyped Individual: Assuming biopax-level2:id699451362_Smad7 is an
individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id1657654683_id1285092580 is an individual"
"Untyped Individual: Assuming biopax-
level2:id687446721_Transcription0 is an individual"
"Untyped Individual: Assuming biopax-
level2:Connected_id121677340_id933799551 is an individual"
Regards,
Alan
Ken Fukuda
#Sorry for my late reply. I got a fever and wasn't in
#my office for the latter half of last week.
Thanks for your review. I will check what happened with
things you pointed out below.
Here are the latest version for today.
[Modifications]:
No empty SequenceLocation instances.
No direct assertion in PARTICIPANT property
There were some pathway instances without evidence instances.
There was one error in the original data (using CONTROL edge instead
of OUT edge).
Merged duplicated instances of interaction class, pathway class,
physicalEntityParticipant class.
[not done yet]:
name space abbreviation mentioned by Alan Ruttenberg
[known bugs]:
*instance of evidence class fails to escape PubMed in XREF
*Event node's Evidence not mapped to control class and conversion class
[attachment file]:
2006Jun13-TGF-beta_super_family_signaling_pathway_canonical.owl
2006/6/9, Alan Ruttenberg <alanrut...@gmail.com>:
Ken Fukuda
> <bp:sequenceSite rdf:ID="id1657654683_id1709918560_ss">
> <bp:POSITION-STATUS>CERTAIN</bp:POSITION-STATUS>
> <bp:SEQUENCE-POSITION rdf:datatype="http://www.w3.org/2001/
> XMLSchema#unsignedLong"/>
> </bp:sequenceSite>
>
> In this the value CERTAIN isn't one of the choices (EQUAL, LESS-THAN,
> GREATER-THAN)
> Also, the SEQUENCE-POSITION doesn't have a value (it needs one if
> you are listing it. If the value is unknown you would simply not
> state it). If there isn't a SEQUENCE-POSITION that is known, it's not
> clear what the point of having the sequenceSite instance is.
This is a bug. We have a standard BioPAX exporter and an extended
BioPAX exporter
but there were cases where the specified option was not taken.
Partly to solve this bug, we will change the mapping in the next release.
INOH's connected edge was mapped to sequenceFeature and during the
export process sequencePosition instances were also created.
In the next release, we will map INOH's connected edge to
relationshopXref class (like our PASSING edge)
> As I noted in the previous version, PARTICIPANTS is still being
> directly asserted.
Yes. This is solved in the June13 release.
> In the following
>
> <sequenceParticipant rdf:ID="id1657872740_MAPK0">
> <ORGANISM rdf:resource="#Canonical"/>
> <PHYSICAL-ENTITY rdf:resource="#MI0003570_MAPK"/>
> <CELLULAR-LOCATION
> rdf:resource="#plasma_membrane__extrinsic_to_internal_side_of_plasma_mem
> brane_"/>
> <EVIDENCE rdf:resource="#id1657872740_MAPK_evd1"/>
> <XREF rdf:resource="#INOH_II0000028_id1657872740"/>
> </sequenceParticipant>
>
> ORGANISM is a property of rna, protein, pathway, complex, or dna, but
> not participants.
> EVIDENCE is a property on interactions or pathways, but not
> participants
> similar issue with XREF, which isn't valid on participants
You are right. Again, this is a bug. The standard/extended BioPAX
switch was not working correctly. We are fixing this right now.
This is a bug and I believe it has been solved in the June13 release.
In the next release, INOH connected edge will be mapped to BioPAX's
relationXref class.
Thanks for reviewing.
Best,
Ken
Ken Fukuda
The latest INOH owl file.
Release note (version. 06/19/2006):
1.sequenceSite has "CERTAIN" value
Before: when converting connected edge to sequenceFeature, we also
created sequenceSite instance.
After: convert connected edge into relationshipXref instance and do not create
a sequenceSite instance.
2.changes in DB, ID, COMMNET in unificationXref instances.
DB：
(Before) IEV_0000005 -> (After) http://www.inoh.org(EventOntology)
ID:
(Before) Phosphorylation -> (After) IEV_0000005
COMMENT
(Before) http://www.inoh.org(EventOntology) -> (After) Phosphorylation
3.declare xsd ( xsd="http://www.w3.org/2001/XMLSchema#") in the begining
and abbreviate URI description.
2006/6/17, Ken Fukuda <fukud...@aist.go.jp>: