Herbert Sauro wrote:
> And don't fall into the trap of spending
> 6 weeks trying to work out why a particular simulation did a certain thing
> when an electrical engineer sitting next to me pointed out to me what was
> going on in about 2 minutes flat. Trouble with large models is that one
> can get blinded by all the detail.
Indeed, but the real opportunity now with so much whole-genome data
circulating around is to create such a large model just as a starting point
and then to simplify it to the point of making it understandable (still
keeping its predictive ability). This is where modeling becomes
interesting!
The main advantage of whole-genome data is that the observation is not
biased (i.e. we are looking at all components of the system); another
positive aspect is that the data are consistent (i.e. all the components
measured in the same conditions). The latter point is the main difficulty
in creating bottom-up models where the kinetics of each enzyme are measured
in different conditions often making the models inconsistent.
--
Pedro Mendes
Virginia Bioinformatics Institute, Virginia Tech
1750 Kraft Drive, suite 1100, Blacksburg, VA 24060, USA
Phone: +1 540 2317411 Fax: +1 540 2311777
