Biomedical Engineering Reference
In-Depth Information
In this trajectory, the assumptions and limitations are explicit and well-controlled
hypotheses can be formulated. The next two trajectories are retrospective. First,
there is a growing field adding thermodynamic complexity to phylogenetic models
and this will be described. Second, another growing field involves the use of
standard population genetic models for interspecific evolution, also incorporating
increasing biophysical reality in parameterizing the models. These will be described
in turn.
2
Simulation and Forward Evolution
Simulation and comparative genomic analysis represent complementary trajectories
in evolutionary analysis. As events become more ancient, reconstruction of evolu-
tionary histories and substitutions underlying the sequence-structure-function link
become harder to recover. Further, interpretation of underlying mechanisms and
controlling data for numerous potential variables become potentially problematic.
Therefore, simulations of evolutionary processes with different mechanisms and
explicit assumptions can provide insights into evolutionary pathways that are
difficult to obtain from comparative genomic data.
Simulation approaches have a long history in the field of evolutionary biology,
for example, in the context of population genetics [
1
]. Calculations on a population-
wide level of variation in allele frequencies due to different fitness effects are not
computationally demanding and provide insights into population-level processes.
However, because of the lack of biochemical detail in such models, they tell us
very little about the mechanisms that underlie fitness changes. To model the effects
of mutations on molecular evolution, it is necessary to have at least a minimal
representation of protein structure.
The most basic such model is the hydrophobic-polar lattice model [
2
]. The
protein is represented as a series of interconnected beads on a rectangular grid, in
either two or three dimensions, each of which represents either a hydrophobic or
a polar residue (Fig.
1
a). Hydrophobic residues interact favorably, mimicking the
solvation pressure to form a hydrophobic core, and other interactions are typically
neutral or repulsive. This simple model enables sampling a very large number of
configurations rapidly, and in the two-dimensional case it even allows a complete
enumeration and examination of the entire sequence-structure-fitness landscape.
Bornberg-Bauer and Chan [
2
] used this technique to examine the distribution of
thermodynamic stability for sequences undergoing neutral evolution, and found that
the funnel-like behavior of the energy landscape of protein folding is recapitulated
by the sequence landscape. Melin and co-workers [
3
] found two criteria that apply to
all protein-like sequences under such a model. The native state is highly designable
(robust to mutation) and is well separated on the energy landscape from random
configurations. These properties are similar to those observed in real proteins.
The next level of complexity involved the usage of the 20 amino acids in lattice
simulations. Sali and co-workers [
4
] examined the folding process using a 3D
Search WWH ::
Custom Search