Biomedical Engineering Reference
In-Depth Information
experiment because of its small genome size (580 kb). The information on the
kinetic properties of genes and proteins was mostly obtained from the KEGG
and BioCyc (previously called EcoCyc) databases.
4.
FUTURE WORK AND ITS RELEVANCE TO BIOMEDICINE
In-silico biological modeling has finally arrived and is here to stay. At one
end of the spectrum, scientists add components and pathways to the system to
enhance its properties, while at the other end cell parts and pathways are re-
moved to rid the system of undesirable components. An optimum way to test the
result of such combinations is to perform computer simulations. The computer
simulations are reasonably accurate, low-cost, fast, and scalable. Of all the mod-
eling approaches, analytical modeling is the most powerful approach, as it
makes it possible to understand the exact regulatory topology of a biochemical
pathway. The best estimate about kinetic parameters can be made either directly
using standard experimental protocols or by inference from the literature (41).
Living systems represent a continuous and nonlinear interaction of sub-
stances that are not only large numerically but divergent in variety. Due to the
inherently dynamic nature of a biological system, the traditional modular ap-
proach does not hold good in all situations. What is needed is a systems-based
approach that not only models the current state accurately but also predicts all
possible future states in the presence of varying environmental conditions and
perturbations. The systems approach offers a possibility of addressing such
questions as follows: Given a certain gene mutation, what would be its down-
stream impact on the immediate and/or related regulatory and metabolic path-
ways? How many parameters (and in what combinations) can be tweaked to
produce a continuously dividing malignant cell? Given a good quantitative
model of a parasitic metabolic pathway, which are the most important and
highly connected nodes that can be perturbed to produce large-scale effects?
Assuming that a drug binds specific proteins or genes, how many pathways will
be affected, and in what sequence, both upstream and downstream? Thus, it is
even possible to determine the "virtual side effects" of a drug by conducting
such computer-based experiments. However, to make such virtual experiments
more accurate, noise-free high-throughput data coupled with a reliable in-vivo
validation system is required. The future will see the significant growth of inte-
grative models that not only consider different cellular processes in parallel (me-
tabolism, gene expression, and signal transduction) but also combine diverse
modeling strategies (deterministic and stochastic) (30,33).
One of the problems with porting in-vitro data to computer models is the
gross dissimilarity between in-vitro and in-vivo systems. While the former are
buffer based, the latter represent gel-based environments. Thus, we need better
assay systems that provide conditions similar to in-vivo situations. In the future,
