We use genomic strategies  to investigate the role of genetic and epigenetic variation in the regulatory domain of the genome, with an interest in both human disease and evolution. The two main areas of interest of the lab are:

• The evolutionary biology and potential contribution to human disease of components of the epigenome.
• Annotating the regulatory elements of the human genome using a combination of comparative analysis of sequenced vertebrate genomes and public  epigenomic datasets.

We have developed a method that uses high throughput second generation sequencing to survey DNA methylation states on a genomewide scale. The method is sufficiently inexpensive to be applied to large number of samples and used in comparative studies. We call this approach “methyltyping”, since it provides the epigenetic field with the same ease of assaying epigenetic states as genotyping does in the genetic field. Our phylogenetic comparison of methylation state differences in multiple  humans, chimpanzees and orangutans provides evidence for epigenetic changes that occur in the germline and distinguish closely related species, and suggest that germline epigenetic states might constrain somatic states. It also revealed a list of regions whose methylation state has changed during human evolution and that may have played a role in the divergence of human and chimp. The methyltyping approach is perfectly suited for investigating the role of variation of methylation states in human disease; we are expanding this approach to study variation in DNA methylation states in human populations and its possible correlation to disease risk.
We have used a different methylation mapping method, also based on high throughput sequencing, to solve a long-standing puzzle: the presence of DNA methylation in the genome of the fruitfly Drosophila melanogaster, the most common model in animal genetics. We find that the fruitfly genome is indeed methylated, but in a pattern that is very different from that found in vertebrates. This finding opens the opportunity to characterize the role of methylation during drosophila development.
We also use deep sequencing to investigate the function of piRNAs, a class of very abundant short RNAs that may participate in establishing the sites of DNA methylation in the genome. We find that piRNAs and their PIWI protein partners are expressed in somatic tissues (and not only in the germline, as it is commonly assumed). The presence of piRNAs in cell types amenable to culture allows us to manipulate components of the piRNA pathway to obtain insights into how the pathway operates.
Finally, understanding the role that DNA sequence variation plays in human disease requires the ability to predict the functional impact of individual sequence variants. In collaboration with the lab of Inna Dubchak at the Lawrence Berkeley Lab, we use comparisons of genomes of multiples sets of related vertebrate species to identify sequences that may play a functional role within these species.  We also leverage the deep datasets of human sequence variation generated by the 1,000 Genomes Project: the patterns of rare and common sequence variants allow us to tease out regions of the genome with higher likelihood of being functional; we then combine the results of this analysis with public epigenomic datasets to obtain a functional annotation of the noncoding portion of the genome.

Revised: Wednesday, October 10, 2012 9:42 AM