Research Program

Hidden in the features of Nature's creations is a living-record of the great unfolding of life on Earth. Despite their different uses, the limbs of bats, birds, humans, and even whales are built from the same constitutive parts. The 'same' features in different organisms despite their different uses are called homologies. While homology is well known as the basis of taxonomic classification, its existence on a larger scale is manifested as the variations-on-a-theme diversity readily apparent in animals, plants and fungi. This regular pattern in organisms due to common descent makes it possible to study animals and plants to learn about human biology.

At the same time, different parts of animals - and hence different parts of their genomes - evolve at radically different rates. Some genes evolve extremely quickly, for example, genes with immune-related functions (Castillo-Davis and Kulathinal et al. 2004). Other genes evolve extremely slowly or not at all, for example genes involved in basic cellular functions like protein synthesis and mRNA splicing (Castillo-Davis and Kulathinal et al. 2004). Gene duplication, cis-regulatory evolution, and even intron gain and loss appear to behave equally capriciously in their evolutionary dynamics (Castillo-Davis and Hartl 2002; Castillo-Davis et al. 2004b; Castillo-Davis, Bedford, Hartl 2004c).

Since the rate of change of different animal features is so heterogeneous and seemingly complex, it would be useful to discover a set of general genetic rules to explain and predict why particular morphological, physiological, and molecular features are conserved between species whereas others are not. The discovery of evolutionary genetic rules that result in "phenotypic inertia" is a central research goal of our lab.

Current Research

Noncoding cis-Regulatory Sequence Evolution

Illustration of the shared motif method (SMM)

Differences in gene expression between species may entail changes in spatial, temporal, and environmental dimensions. The genetic basis of such changes and how they evolve is currently not known. By identifying the types of cis-regulatory changes that affect quantitative changes in gene expression between species we aim to elucidate some of the general rules that underlie gene regulation. Toward this end we combine different types of genomic expression and sequence data to model how cis-regulatory sequence change and gene expression are related.

Gene Network Evolution

Networks of interacting genes are responsible for generating life's diversity and for mediating how organisms respond to their environment. Gene networks play a central role in our sophisticated immune response, the ability to digest food, and even for causing cancer - the disaster that occurs when gene networks become un-regulated. Thus understanding the properties of gene networks is of fundamental importance in "post-genomic" biology.

However, very little is known about already well-characterized genetic networks. Outstanding and fundamental questions include: How much natural variation is present in gene networks among individuals? Among species? What is the genetic basis for such differences? Do certain parts of gene networks change their ability to tolerate variation (mutational insult) over evolutionary time?

The comparison of a well-understood gene network among related species is a powerful way to gain insight into these questions. Each species represents a natural experiment and, when compared, can shed light on the structure, dynamics, and function of a gene network. Our goal is ambitious: to discover general principles underlying gene network function, variation and evolution.

To do this, we use a combination of statistical and computational techniques coupled with molecular genetic experiments in the fruit fly, Drosophila melanogaster and its relatives. The fruit fly is our chosen organism of study because of the powerful molecular and genetic tools available to manipulate and monitor this organism. The genomes of 11 other species in the genus Drosophila have recently been sequenced. These species span a range of 3 million to 40 million years in relationship to one another. The ability to compare the genomes and phenotypes of multiple fly species provides an unparalleled resource that we will use to determine how gene networks function and evolve.

Past Research

Software and Algorithm development for genomic/proteomic analysis

Comparison of average intron lengths and total exon lengths in genes with high and low expression (genes from H. sapiens).

Mixed-effect model representation of gene expression over time.

Estimated mean expression curves and 95% confidence bands for four of 17 clusters discovered by SSC in the D.melanogaster time course microarray data.

Duplicate genes exhibit accelerated rates of intron gain/loss in comparison with orthologous genes at almost all levels of synonymous divergence (dS).

Biology has gone from being a data-poor science to a data-rich one and thus presents an exciting challenge not only for biologists but also statisticians, computer scientists and other quantitative workers. The ever-growing bounty of - omic data (genomic, proteomic, and phenotypic, to name but a few) generated by technologies enabled by genome sequencing projects is quickly out-pacing our ability to meaningfully synthesize it.

Thus the development of tools for post genomic analysis - the interpretation and synthesis of thousands of data points from a chemical, clinical, evolutionary, or other perspective - is an important aspect of our research program. Toward this end we have developed and published software tools for comparative genomic sequence analysis (Shared Motif Method), functional genomics analysis (GeneMerge) and gene clustering for time-course microarray data (SSClust). Please see Software more information on these applications.

High-throughput phenotyping & fast-mapping of QTL using tiling microarrays

Expression level of the genes of the phototransduction pathway in D. melanogaster and D. simulans are largely conserved between the two sibling species.

coming soon...

Development of theory and tests of homology and character stasis

coming soon...

Species delimitation in the fossil record using extant intra- and inter-species variation

Scatterplot of dN versus dS for pairwise mammalian orthologs.

coming soon...