This page is under constructinon 
Hopefully this page will evolve into a set of semantically integrated notes representing the BioPAX Boston Meeting July 1, 2005 3PM-6:15PM rather than the set of individual notes below. Those who haven't contributed their notes yet, please aggregate or integrate them into the section below.
Joanne's questions for Tom Atwood:
Tom (I think) mentioned something about Jike and Jude? - there was pressure in a domain for the project to go open
SIMILE - digital media Library Creative Commons / Science Commons - digital rights in a flexible format use limits digital signal (whatever that meant).
can you protect what your putting out? Now that pharma wants to point it out. Resctricion - science commons thinking
Tom asked if there were probablistic models pathways.
******** = Joanne Luciano's Notes on 0
0)Discussion of ontology
JSL -mention barry's Ismb paper (note 2 bugs)
mention obo "relationship ontology" which biopax should adopt
note that part of and transforms are most important and biopax needs to have
"transforms may have application to states"
Pathway visualizatoin, representation, Barry Smith - how to put solid meaning on text relationships in pathways.
when you look at a statement, you don't know what it means.
"Not theory neutral and not theory"
Getting core of interaction right
GO relationship between classes - can be instance of as well. paper: is-a aint (tried to find the paper, couldn't) the suggestion was made to abandon 'is-a'
******** Dan Corwin's Notes on 1
1) Proposal C on "States" took up most of time, and generated some good, philosophical comments on what seemed needed. Observations I recall were:
(post-mtg) "States" Yahoo-group Word docs as background would have made this review much easier. Instead, we tried to infer why "C" was actually cast as it appeared.
No explicit States requirements or proposal selection criteria seem defined but for "simplicity" and "backward compatibity" being good. That needs fixing.
A suggested criteria: insist a proposal easily handle an expected stream of new "states" being identified, or old ones redefined. (Implies high modularity)
A linguistic analogy: in the BioPAX language, interactions are like verbs, participants are like nouns. Level 3 now adds states as "modifiers".
Why is "State" not just the root of a taxonomy of Property classes, in which domain and range of each node says what is modified and the state options? We discussed this at length.
When does such a state Property need to be more than a DataTypeProperty? Possibly when its value must model (as an object) specific events that led to the state.
******** = Joanne Luciano's Notes on 1
Ontology discussion Reaction + Participant
Participant in state
What is the goal of representing states? Record protein-protein interaction + cellular location, sequence
Why do you want extr? objects want to have physical entity intended for re-use in multiple reactions
ATP is not phosphorylated ADP
Protein with tweakable domains
Salted vs Unsalted - states for small molecules
Cellular location - compoents ebmedding
+
> applies to everything to be part of PhysicalEntity can keep it anonymous, or name it.
refer to state independent of PEP
GOAL: new type of state and want to add it
open question
whether any aspect of state information will be reuasble
when we have use cases then we should make an object out of it
typially have partial, we need to have a way to state it and wen you konow more, a migration path to add knowledge
openworld object and state are both possible
Ulrich presentation, second day of BPC SIG Piere Borg
PEP glucse transport as cell location psysical entity participant glucose gluc local peoperty of ?? means "some process like this"
[stateFor GSK3; phosSer25; ...] [stateFor GSK3; location; membrane]
[drawing of ontology]
state: a short hand for knowing something is true versus knowing how it came to be true
Process Modeling"
Pi Calclus Ambient
predicate calculus + time and location
XVPI splice and unspliced form, shows up as 2 bands, not alternative spliced, spliced later
No meaning with states add meaning by adding processes
by class by ??
Know phosphorylates
canditate - phosphorylate as a prototype
or if known, point specifically to ...
has to admit state transition physical entity in same state
atomic pathway model left hand side right hand side extreme case of state change (diff models) point change
ex. tanscription complex, it has set of states, in each state has states within each state processes interact possibly addressed in "steps" assembling to fully functioning
3 contexts
cellular context (compartment; location) cell cycle time type of cell
******** Dan Corwin's Notes on 2
2) Eric mentioned a need for named subgraphs and emerging tools which would let them be managed and modeled as cohesive groups of RDF assertions (not unlike Conceptual Graphs).
I said the WORDS Modeler (or any TopicMap) can do this now using the Scope mechanism (to bundle RDF assertions into groups tagged by any set of topics - such as author, version .. )
Nada's paper is now in the wiki. Dan's topic map engine and dumps it out (not yet RDF; summer student was to do with).
An OWL reasoner could NOT handle "Scope" today, but the feauture could help one query a context to export assertion-groups back into RDF - which today's reasoner could use normally.
(post-mtg:) an OWL-DL model of topic maps is due in August. That and a bit more time for RDF-XTM convergance should let OWL reasoners hack XTM as easily as RDF.
= Joanne Luciano's Notes on 2
How do you passively view or contribute?
Named Graphs:
G1: slightly different from G2:
x y z x y z x w c x w k z y k z y x
they don't get merged.
=========
3) Discission of BioPathways Consortim Role in the Communty
Eric Neumann spoke of the BioPathways Consortium looking to produce a proof of conept, support large scale community effort through prototyping, e.g. semanitc web, constructing views of scienceand provide published demonstrations. He talked about a potential collaboration between MIT and Stanford.
Put up public resources allow others to read/annotate -S-wiki
piggybank plugin - tag & look at RDF structures if you have an account at similie givng peple access to wikis
put them up in a way that allows people to work with the tools and concepts to generate pathway - have people point to each other/semantic blog
Possible callob between MIT and Stanford (Eric spoke with mark musen)
Eric mentioned the YeastHub a paper presented at ISMB as one to look at.
take BioPAX into an annoted model.
Eric wanted to know if an anntotated BioPAX model were set up, would current projects such as those at Millennium or elsewhere, be willing to work with them?
Jonathan said, not at Millennium. They would have to make an investment in RDF. They already have non-RDF tools. However, if the tools were interconvertable, then they would.
Alan Ruttenberg said instead of Merging 2 biopax sources that talk about the same thing, each could become an annotation of the other.
Is it useful from a research perspective (for Millennium?)
Jena 2.0 can handle multimodels
What projects would be willing?
Dan Corwin: Topic Maps do this with Scope. However, his modeler does not yet write out...
N-tuples rather than triples import + BioPAX in common document
Some notions for how the BPC could expand its offering to more than just one event per year at ISMB.
Publish RSS Events other than at ISCB
- What would folks like to see from BPC as a resource?
Website:
Local meetings (Suggested by Jonathan, Joanne thinking - Science & the city) One place for related meetings
Aggregator of pathways related projects
User groups for BioPAX (which is broader and different from the standards group)
Post query for semantic geography --
4) Discission - test and validate - not discucssed here, but mentioned it was discussed at ISMB -Frank Gibbons and Joanne long discussion about mutliple ways to approach testing of BioPAX ontology and data.
= Allen Ruttenberg's Notes
= Joanne's notes of things Eric Neumann said. Eric, please flesh these out and incorporate them in the above. Thanks.
5) Discussed Presentation From PSB-2001 on Molecular Interaction Graphs
Jonathan Rees' Summary Here:
NOTE: BRING OR GET WHITE BOARD MARKERS FOR NEXT BostonBioPAX meeting.