Using DNA sequence evolution to locate gene regulatory elements

The human genome is 3.4 billion letters long; of this, we can confidently ascribe a function to perhaps 1%, the protein coding complement of our genetic material. The function of other the 99% is still largely unknown although we know there must be sequences which regulate where and when genes are turned on and off. However, locating these regulatory sequences has proven very difficult. Current methods are at best only 50% successful and may miss many of the most regulatory elements, those that have undergone evolutionary changes. We plan to investigate a new method which may help us locate some of these regulatory sequences which have evolved. The method simply looks for sequences whose rate of evolution correlates to rates of gene expression evolution. We plan to investigate the method in primates, fruit flies and yeast.

One of the great challenges of genomic science is understanding the function of non-coding DNA which may comprise over 99% of a genome. It is known that in this large amount of non-coding DNA there are sequences which regulate gene expression but these sequences have proven very difficult to locate. Currently the best method for identifying regulatory elements on a genomic scale is phylogenetic footprinting or shadowing. However, this method is at best only 50% successful in identifying gene regulatory elements and the elements which are identified are typically those highly conserved by evolution. In this project we plan to investigate a new method for identifying regulatory elements which subverts the logic of phylogenetic footprinting and rather than search for regions which are highly conserved, we search for regions whose rate of evolution correlates to the rate of gene expression divergence. We will investigate this method in three species groups: primates, Drosophila and yeast. Our aim is simply to investigate whether there are correlations between sequence and expression divergence, and if there are, to quantify how common these correlations are. If there does appear to be a signal we aim to apply for further funds to test the putative regulatory elements for regulatory activity.