> Mike Cantor asked:
> >
> >2. What are the prospects for larger-scale modeling projects like the
> >NIGMS's projected 25$ glue grant for a "Virtual Cell." Is this a naive and
> >premature waste of money, or will it be a fountain of knowledge.
>> I'm very sceptical about this, but open-minded enough to be ready to be
> convinced otherwise by hard evidence. The most complete metabolic model I
> know of is the model of glycolysis in Trypanosoma brucei worked out by
> Barbara Bakker. Even that only includes around 60% of the numerical values
> one would ideally like to include (the others having to be guessed, treated
> as "negligible", etc.), but it still requires a tremendous amount of
> high-quality kinetic information about the component enzymes. When I
> counted several months ago I think (I'm trying to remember, not checking!)
> there were about 65 separate pieces of numerical information in the model,
> each of which was the result of some serious experimental effort. In the
> case of whole cell models I don't believe we are anywhere close to have a
> comparable amount of genuine information to feed in. I'm not ready to put
> much confidence in the results from a model where few of the numbers used
> are based on experimental kinetic information.
Actually I am very favourably disposed to the e-cell type of project,
and I am hopnig that the UK can start one (it will also bring in all
the mol biol people who want to get into /metabolic modelling but
don't know the right people to hold their hands).
The point about not having much (and certainly not ENOUGH) good
kinetic data is well taken, but I'd say the following:
(i) what modelling CAN do is allow one to look at the SENSITIVITY of
what one is interested in (e.g. a specific flux or subset of fluxes)
to different parameters. In practice I suspect that most parameters
will have hardly any effect on many of the variables over a wide
range, so having wrong values there doesn't hurt too much. The model
will tell you which ones you HAVE to get right (or rightish)
(Comment: the next generation of metabolic models, which will
probably be integrated with genomics databases, would do well to
include fields containing information on the EVIDENCE for particular
values, see e.g. Evidence Codes in Gene Ontology
http://www.geneontology.org/GO.evidence.html).
(ii) the availability of lots of omics methods (transcriptome,
proteome, metabolome) will provide HUGE datsets which will make it
MUCH easier than in the past to constrain our models, so this too
gives me cause for optimism, and if we are going to have DATA on all
or most genes/proteins/metabolites then we need to have MODELS on all
or most genes/proteins/metabolites.
(iii) There is a view that says that kinetics as measured in vitro or
even in situ are so different from those occuring under the actual
environmental conditions in vivo that they should not be treated as
much better than estimates anyway, and that that will NEVER change.
In which case we shouldn't worry about dotting the last i, etc.
(iv) I think there are larger models than Barbara's trypanosome one,
though I am not sure how much they are based on hard experimental
kinetic evidence. I think Bernard Palsson has some bigger ones but
teh paper is not to hand. (I remember a poster at Il Ciocco 1989 by a
guy from I think New Mexico that had SIX HUNDRED parameters for an
erythrocyte model....never heard of it again).
Happy 2001.
Douglas.
********************************************************************
*(Prof.) Douglas B. Kell Phone: +44 1970 622334 *
*Director of Research *
*Institute of Biological Sciences Fax: +44 1970 622354 *
*Cledwyn Building *
*University of Wales, Email: dbk at aber.ac.uk *
*Aberystwyth SY23 3DD, U.K. http://gepasi.dbs.aber.ac.uk/ *
********************************************************************
********************************************************************
---
