Thomas Butler

Office: 3-105 ESB
Department of Physics
University of Illinois at Urbana-Champaign
1110 W. Green St.
Urbana, IL 61801-3080
USA

email: tbutler2 [at] illinois [dot] edu

About Me

Research Interests:

My primary interest is theoretical statistical physics. I have applied the tools of statistical physics to biological systems to develop novel data analysis methods as well as address problems in ecology and population dynamics. My work is advised by Professor Nigel Goldenfeld. More information about the research interests of the group members can be found through the group's website.

Evolution and Optimality in the Genetic Code

Studying the kinds of optimality present in the genetic code can provide evidence of the kind of evolutionary dynamics responsible for its origin and structure. Using methods inspired by statistical physics, my collaborators and I have uncovered evidence that the genetic code possesses extreme optimality with respect to absorbing the effects of point mutations. We have also shown evidence that the code contains signatures of a stage in its evolution dominated by "statistical proteins" – proteins that are defined by equivalence classes of similar amino acid sequences rather than by unique sequences. Both of these discoveries point to collective evolution of the genetic code through pervasive horizontal gene transfer. This work was recently published as Rapid Communication in Physical Review E. A preprint is here. We've also applied these techniques to analyze a specific proposed precursor to the genetic code. This work will be published in Physical Review E as a Brief Report and is available as a preprint here.

Nonlinear Stochastic Processes in Population Dynamics

Methods adapted from quantum field theory have proven useful for systematically treating the effects of noise intrinsic to individual based models of population dynamics as well as the nonlinearities inherent in any interacting population. In the case of classical Lotka-Volterra predator-prey models, these factors lead to large beyond mean field stochastic oscillations in time, though not in space. We have derived an analytical expression for the power spectrum of these oscillations. This work was published as a brief report in Physical Review E. It is also posted on the arXiv.

We have also recently applied the same techniques described above to analyze a model of predator-prey interactions that contains spatial pattern formation for some parameters when formulated as a partial differential equation. We showed that when this model is analyzed as an individual based model, patterns form over a much larger set of ecologically interesting parameters than predicted by the partial differential equation formulation. We also were able to make specific predictions about the power spectra of real ecosystems in which spatial pattern formation is driven by noise. This work has been published in Physical Review E as a Rapid Communication. A preprint of this work is posted on the arXiv.

To develop intution for the spatiotemporal dynamics of the pattern forming ecosystem described above, we have developed a simple agent based model using the Netlogo modelling language. This model can be explored here. Please be patient while Java loads.

Arms Control

When time allows, I also maintain an active interest in arms control. In particular, I have been interested in nuclear weapons policy issues where an independent, civilian, assessment of the technical issues can inform key policy decisions. This work is carried out in affiliation with the Institute on Global Cooperation and Conflict.