The aim of our research is to understand phenotypic evolution by studying the processes by which the genetic networks underlying development diverge. A major feature of developmental networks is their robustness (Masel & Siegal 2009). That is, they are tolerant of both environmental and genetic perturbations. Our lab uses both experimental and computational approaches to understand the causes and evolutionary consequences of this robustness.
One major experimental focus in our lab is on the process of sexual differentiation in Drosophila melanogaster and related flies. Sexual differentiation is a powerful model system for studying the evolution of development because many aspects of sexual morphology, physiology and behavior differ between closely related species, thereby enabling high-resolution comparative analysis. Despite this rapid divergence, sexual traits are highly robust within species and indeed are often diagnostic of species. We are using genomic and genetic approaches to identify and characterize regulatory pathways involved in genital development and function in D. melanogaster (e.g., Chatterjee et al. 2011). This will lay the groundwork for determining how these robust pathways diverged between closely related species.
A second major experimental focus is on directly identifying and characterizing genes that contribute to robustness of many traits. We have screened the genome of the laboratory yeast, Saccharomyces cerevisiae, for genes whose deletion increases the variation in the morphologies of individual, genetically identical cells. Yeast is advantageous for this work because of its wealth of genetic and genomic resources, and because it lends itself to high-throughput analyses. Hundreds of yeast genes increase morphological variation when deleted, and these genes tend to be highly connected in cellular networks (Levy & Siegal 2008). We are currently testing whether the same genes also buffer genetic differences between cells, and whether the variation that is revealed by impairment of these genes is potentially beneficial. This work is complemented by theoretical investigations into the evolution of complex gene networks (e.g., Bergman & Siegal 2003; Siegal et al. 2007), as well as bioinformatic analyses of regulatory networks (e.g., Chen et al. 2010), which give us predictions to test experimentally.