Next Meeting Tomorrow
Oliver Ruebenacker
Our next meeting is tomorrow, Tuesday, April 14, 4pm to 6pm, in 346
Stata, MIT, and on Skype.
What would you like to see on the agenda?
Alan, Michel, last time you said you wanted to focus on pathways
rather than models, or something like that. Can you specify what you
had in mind?
Take care
Oliver
--
Oliver Ruebenacker, Computational Cell Biologist
BioPAX Integration at Virtual Cell (http://vcell.org/biopax)
Center for Cell Analysis and Modeling
http://www.oliver.curiousworld.org
Jonathan Rees
On Apr 13, 2009, at 9:30 AM, Oliver Ruebenacker wrote:
>
> Hello, All,
>
> Our next meeting is tomorrow, Tuesday, April 14, 4pm to 6pm, in 346
> Stata, MIT, and on Skype.
>
> What would you like to see on the agenda?
>
> Alan, Michel, last time you said you wanted to focus on pathways
> rather than models, or something like that. Can you specify what you
> had in mind?
if I may hazard:
Anything can be a model of anything else; you just need a way to
interpret the elements of the one thing in as elements of other thing.
So a logical theory can be interpreted as bits of reality (if you have
a mapping of the theory's terms to things in reality); the
interpretation is then a model of the theory if relationships are
carried over (e.g. if 'x a member of C' in the theory means that x
(the interpretation of 'x') is a member of C (the interpretation of
'C') in reality). Of course an interpretation may not be a model of
the theory at all, if relationships aren't carried; and the theory can
have many other models. (This is called 'model theory' in logic.)
Contrariwise, reality can be interpreted (encoded, curated) in a
logical theory. If theorems of the theory are interpretations of true
relationships in the world, then the theory is a model of reality.
Popper tells us that we'll never know whether any theory containing a
universal quantifier (class - think DL) corresponds to a true state of
affairs in reality. This doesn't mean it's pointless to build
theories; it just says we should try to make theories that *to the
best of our knowledge* are true - that is, are reasonable candidates
for being models of reality. We can generally call them "models"
instead of "putative models" without getting confused because the
"putative" goes without saying.
I think the sentiment being expressed is that when you "do ontology",
you implicitly construct a logical theory that you hope is a model, by
virtue of the way you decide to talk about things (the classes and
relationships you choose to define). If instead you start talking
about models and their properties and relationships, then you are
implicitly constructing models of models, which is not what we want to
do; we want models of biology, not models of models. And we have a
method for modeling, namely the foundry approach, so we know how to do
it.
Now, fields like systems biology and climate modeling turn the models
(theories, technically speaking) themselves into objects of interest.
That's fine - you can talk about relationships between models,
numerical properties of the models such as convergence, and so on,
intrinsic properties that don't necessarily relate the model to
anything that it's intended to model. The systems themselves can take
on a life of their own and continue to be called "models" even when
they're not models of anything real. That's OK. But I think it's being
suggested is that modeling reality (pathways) should be higher on the
agenda than modeling models.
Jonathan
Barry Smith
>On Apr 13, 2009, at 9:30 AM, Oliver Ruebenacker wrote:
>
> >
> > Hello, All,
> >
> > Our next meeting is tomorrow, Tuesday, April 14, 4pm to 6pm, in 346
> > Stata, MIT, and on Skype.
> >
> > What would you like to see on the agenda?
> >
> > Alan, Michel, last time you said you wanted to focus on pathways
> > rather than models, or something like that. Can you specify what you
> > had in mind?
>
>if I may hazard:
>
>Anything can be a model of anything else; you just need a way to
>interpret the elements of the one thing in as elements of other thing.
you also need a way of interpreting the relations so that they remain coherent
>So a logical theory can be interpreted as bits of reality (if you have
>a mapping of the theory's terms to things in reality); the
>interpretation is then a model of the theory if relationships are
>carried over (e.g. if 'x a member of C' in the theory means that x
>(the interpretation of 'x') is a member of C (the interpretation of
>'C') in reality).
But then 'C' will have to be interpreted as a set (something that has
members), and sets are not things in reality (at least not in the way
that chickens are things in reality). Indeed I believe that there is
no interesting theory all of whose terms can be interpreted as
chicken-like things.
BS
Michel Dumontier
JSL913
level?
What is the boundary of a chicken? At the molecular level?
How many genes in the DNA of a chicken need to be different in order
to characterize it as chicken-like rather than chicken?
> Indeed I believe that there is
> no interesting theory all of whose terms can be interpreted as
> chicken-like things.
Joanne
>
> BS
>
>
>
> >
Alan Ruttenberg
-Alan
Oliver Ruebenacker
Let's postpone or cancel, then.
How about next week?
Take care
Oliver
Andrea Splendiani
same for me & tomorrow.
Next week is ok for me.
ciao,
Andrea
Il giorno 14/apr/09, alle ore 13:25, Oliver Ruebenacker ha scritto:
JSL913
traveling in time and space.... ;-)
J.
Barry Smith
As to boundaries, sets always have determinate boundaries, albeit
determined only down to the level of whole members. (Whether x is a
member of the set y is always a black-and-white issue.) Sets on
whatever level are radically distinct from chickens in the following
respects: sets do not exist in space and time; sets have members;
sets have subsets.
BS