BIOLOGICAL INSPIRATION FOR COMPUTING 291
8.4.2.4 Applications
While significant from a research view, synthetic biology also has practical applications. A strong
driver of this is the rapidly falling cost of custom DNA synthesis. For a few dollars per base pair in 2004,
laboratories can synthesize an arbitrary sequence of DNA;^130 these prices are expected to fall by orders
of magnitude over the next decade. This not only has enabled research into constructing new genes, but
also offers the promise of cost-effective use of synthetic biology for commercial or industrial applica-
tions. Once a new lineage is created, of course, organisms can self-replicate in the appropriate environ-
ment, implying extremely low marginal cost.
Cells can be abstracted as chemical factories controlled by a host of process control computers. If the
programming of these process control computers can be manipulated, or new processes introduced, it
is—in principle—possible to co-opt the functional behavior of cells to perform tasks of engineering or
industrial interest. Natural biology creates cells that are capable of sensing and actuating functions: cells
can generate motion and light, for example, and respond to light or to the presence of chemicals in the
environment. Natural cells also produce a variety of enzymes and proteins with a variety of catalytic
and structural functions. If logic functions can be realized through cellular engineering, cellular com-
puting offers the promise of a seamlessly integrated approach to process control computing.
Synthetic or modified cells could lead to more rational biosynthesis of a variety of useful organic
compounds, including proteins, small molecules, or any substance that is too costly or difficult to
synthesize by ordinary bench chemistry. Some of this is already being done by cloning and gene
transfection (e.g., in yeast, plants, and many organisms), but synthetic biology would allow finer con-
trol, increased accuracy, and the ability to customize such processes in terms of quantity, precise mo-
lecular characteristics, and chemical pathways, even when the desired characteristics are not available
in nature.
8.4.2.5 Challenges
Synthetic biology brings the techniques and metaphor of electronic design to modify biomolecular
networks. However, in many ways, these networks do not behave like electronic networks, and the
nature of biological systems provides a number of challenges for synthetic biology researchers in at-
tempting to build reliable and predictable systems.
A key challenge is the stochastic and noisy nature of biological systems, especially at the molecular
scale. This noise can lead to random variation in the concentration of molecular species; systems that
require a precise concentration will likely work only intermittently. Additionally, as the mechanisms of
synthetic biology are embedded in the genome of living creatures, mutation or imperfect replication can
alter the inserted gene sequences, possibly disabling them or causing them to operate in unforeseen
ways.
Unlike actual electronic systems, the components of biomolecular networks are not connected by
physical wires that direct a signal to a precise location; the many molecules that are the inputs and
outputs of these processes share a physical space and can commingle throughout the cell. It is therefore
difficult to isolate signals and prevent cross-talk, in which signals intended for one recipient are re-
ceived by another. This physical location sharing also means that it is more difficult to control the timing
of the propagation of signals; again, unlike electronics, which typically rely on a clock to precisely
synchronize signals, these biomolecular signals are asynchronous and may arrive at varying speeds.
Finally, the signals may not arrive, or may arrive in an attenuated fashion.^131
(^130) One firm claims to be able to provide DNA sequences as long as 40,000 base pairs. See http://www.blueheronbio.com/
genemaker/synthesis.html. Others suggest that sequences in the 100 base pair range are the longest that can be synthesized
today without significant error in most of the resulting strands.
(^131) R. Weiss, S. Basu, S. Hooshangi, A. Kalmbach, D. Karig, R. Mehreja, and I. Netravali, “Genetic Circuit Building Blocks for
Cellular Computation, Communications, and Signal Processing,” Natural Computing 2:47-84, 2003.