Biomedical Engineering Reference
In-Depth Information
simulations must also allow transitions between various granularities of models.
The modeling environment should allow scaling up of a rough coarse-grained
model (based on qualitative data) to a fine-grained model (based on quantitative
data) without having to discard the previous one.
One would also need better data-mining tools and techniques and smarter
algorithms to find a proper genomic syntax that can be fed into the model with a
fair degree of accuracy. As a result of intensive research in metabolomics, it is
hoped that metabolite analysis will provide a clue to novel gene functions. We
still do not completely understand how cells maintain robustness and stability in
environments fluctuating in terms of ion concentrations, nutrients, pH, and tem-
perature. Modelers now assume an ideal situation that does not consider all these
issues, but as more knowledge accrues models will need to be further con-
strained.
An important question that merits answer is: Do networks exhibit symbiosis
and epistasis? If yes, what are the features that promote such crosstalk? Is this
interaction physiology- or environment-driven? What is the role of redundancy
in the evolution of networks? Given that stochasticity in gene expression is de-
termined by extrinsic and intrinsic factors, how does noise evolve over a period
of time? Does noise have any role in pushing gene expression toward more heu-
ristic solutions? Often we curve-fit the data without considering mechanistic
models that might provide real control parameters for the system. But to reach
that state we need a thoroughly validated model that has failed many times over.
In the future we will see more and more forward-looking modeling approaches,
i.e., fitting the biological system to the model, as opposed to reverse engineering
approaches (fitting the model to the biological system). Other areas that merit
attention are: development of a modeling markup language (15); using a com-
mon theoretical framework for representing biological knowledge; obtaining
validated and time-series high-throughput data; and developing tools capable of
integrating large and complex networks. Despite all this, we still do not know if
mathematics is the right tool for representing biological systems? If not, what is
the best way to model the dynamic cell? Is there a "law of biological complex-
ity" that has roots in physics or chemistry?
5.
REFERENCES
1.
Ashby WR. 1957.
An introduction to cybernetics
. Chapman and Hall, London.
2.
Bertalanffy L. 1973.
General systems theory
. Penguin, Harmondsworth.
3.
Bower JM, Bolouri H. 2001.
Computational modeling of genetic and biochemical networks
.
MIT Press, Cambridge.
4.
Brown PA, Botstein D, 1999. Exploring the new world of the genome with DNA microarrays.
Nature Genet
21
(suppl.):33-37.
5.
De Jong H. 2002. Modeling and simulation of genetic regulatory systems: a literature review.
J
Comp Biol
9
(1):67-103.
