Statistical mechanics of protein
evolution
Alexandre V. Morozov
(Rutgers University)
Understanding non-equilibrium
transport on networks and in other complex media is crucial for quantitative
analysis of numerous biological, physical, social and technological systems, from
protein-protein interactions in biology to the World Wide Web. While transport
on networks with unweighted edges can be described
using scaling and other analytical techniques, few closed-form results are available
for weighted networks, which appear naturally in considering evolutionary
dynamics on fitness landscapes, chemical reactions, and protein folding. I will
describe an efficient recursive approach for studying
random walks on weighted networks, as well as fitness or energy landscapes of
arbitrary complexity. After demonstrating the approach on some simple examples,
I will apply it to the problem of protein evolution. Specifically, I will
investigate how structural coupling between protein folding and binding (the
fact that most proteins can only function when folded) gives rise to
evolutionary coupling between the traits of folding stability and binding
strength, facilitating emergence of evolutionary "spandrels"
(features that appear through adaptation even though the feature itself does
not contribute to the organism's fitness).