Biomedical Engineering Reference
In-Depth Information
has been concerned largely with transformation—from one species into another
with coincident suites of modified adaptive complexes. Mechanistic biology—to
include physiology and cell biology—has focused on the mechanisms underly-
ing robustness of genotype and phenotype. Thus not only are robustness and
evolvability obverse trends in biological system mechanics, they are also repre-
sented by two largely independent research traditions: the historical sciences
relying on comparative data and theory, and the ahistorical sciences relying on
laboratory data and description.
This caricature of our predicament suggests that two quite distinct problems
need to be overcome in order to develop unified theories of biosystems. One is
to establish the utility of evolutionary thinking in mechanistic science, and the
other is to impress the importance of robustness upon evolutionary theory. Such
a project would go some way towards reintroducing the phenotype into evolu-
tionary theory.
Much has been written on the subject of transformation. Population genetics
is concerned with the study of changing gene frequencies through time (26).
Quantitative genetics is concerned with the change in the mean and variance of
phenotypes across generations (14). In neither case has it been possible to ex-
plicitly incorporate detailed mechanistic components of the phenotye into these
models. A recent movement in this direction involves work on the
genotype-to-
phenotype map
and the
representation problem
(62). The genotype-phenotype
map describes the process of development required to decode a genome into a
viable phenotype. The representation problem is concerned with the way in
which the variational properties of the genome are dependent upon the precise
manner in which phenotypes are encoded in genotypes. To put it another way,
are all phenotypes equally accessible from a given genotype configuration, and
if not, does this depend upon the way in which phenotypes are represented in
genetic data structures? Assuming a fixed representation, are there some pheno-
types that are unlikely to ever be realized even in the face of overwhelming se-
lective advantage? If this is so, then these impediments to isotropic adaptive
transformation are likely to be associated with just those mechanisms ensuring
unity of type, stability of genomes across generations, and homeostatic stability
of the phenotype.
One path through the labyrinth of biological robustness is to keep hold of
two of Ariadne's threads: one connected to limits to evolvability and associated
mechanisms limiting variation, and the other to mechanistic inquiries into ho-
meostasis and the regulation of cellular and individual phenotype (17). Robust-
ness thus relates to two critical properties of complex biosystems: the long-term
limits to evolutionary change and the short-term persistence of system function.
Put differently, robustness mechanisms are one of the bridges connecting the
dynamics of ontogeny with the dynamics of phylogeny by limiting phenotypic
variation and also providing some means of exploring alternative genotypes
without compromising the phenotype.
