Dr. Alexander Stewart

EPSRC PhD Plus 2009 Project
: The evolution of mutational and environmental robustness


I am currently a postdoc based at UPenn in the Plotkin Research Group (http://mathbio.sas.upenn.edu/). In addition I continue to collaborate with Andrew Pomiankowski, Rob Seymour and Max Reuter at UCL.

Research Overview

I am interested in understanding how mutations and environmental fluctuations interact to shape genomic evolution. To do this I use a combination of mathematics and computation to construct and explore the behaviour of population genetic models. I combine this abstract approach with a more empirical approach which involves testing the predictions of my models on the growing body of publicly available genomic data from bakers yeast and closely related species.

Environmental Robustness

All environments are noisy, with fluctuations distributed over all timescales. Fluctuations that occur within an individual's somatic lifetime have the potential to alter the individualís expressed phenotype, resulting in reduced fitness. This is the key insight behind Waddington's concept of environmental canalization, developed over half a century ago. Yet the extent to which environmental canalization, or robustness, evolves, and how this impacts evolution over longer time scales, is not fully understood.

I have recently focused on two projects which have looked at the evolution of environmental robustness in two quite different settings. In the first project, together with Joshua Plotkin and Todd Parsons at UPenn, we constructed a model to investigate how environmental fluctuations over short timescales impact the ability of a population to adapt to environmental changes on much longer, evolutionary timescales. We uncovered a surprisingly rich and complex set of behaviours; increasing the amount of noise on short timescales can either help or harm the ability of a population to adapt on evolutionary timescales.

In the second project, together with Max Reuter, Andrew Pomiankowski and Rob Seymour at UCL, we studied how the noise reduction strategies employed by individual genes differ between haploid and diploid organisms. We showed that negative autoregulation, a common noise reduction strategy amongst transcription factors in haploid E. coli, cannot evolve when two copies of an autoregulating gene are present in the same cell. This was used to explain the almost complete absence of negative autoregulation in diploid S. cerevisiae.

Mutational Robustness

The mutation is the basic unit of evolutionary change, yet many mutations are bad, resulting in an individual with reduced fitness. Just as there is pressure to reduce the impact of environmental fluctuations through environmental robustness, so there is pressure to reduce the impact of deleterious mutations through mutational robustness.

I have been interested in the role played by mutational robustness in determining how genes are regulated. Together with Andrew Pomiankowski and Rob Seymour at UCL, we looked at the evolution of protein-protein interactions between pairs of transcription factors as a mechanism to increase the robustness of regulated genes to mutations at transcription factor binding sites. We identified a tradeoff between the advantage of increased mutational robustness and the cost due to unwanted pleiotropic effects when a new protein-protein interaction is gained. This allowed us to make testable predictions about when a protein-protein interaction will become fixed in a population.

Page last modified on 31 jul 11 22:06