I am broadly interested in statistics, algorithms and their application to problems in genetics and genomics.
Below a few ongoing and current projects are listed.
If you have some joint interests which you would like to pursue together, please drop me a note.
Statistical and Human genetics - I study the genetic architectures of complex traits. Recent GWAS have discovered
numerous new associations between genotypes and phentoypes in humans. For most traits, however, the heritability explained by the discovered SNPs remains, however, modest.
I study different explanations for the 'missing heritability'.
In particular, I study the effects of genetic interactions (epistasis), statistical power issues, and perspectives from population genetics theory. I also develop novel methods for estimating heritability.
Comparative Genomics - The avaialability of new copmlete genomes for hundreds of species yields an unpresented oppurtunity to learn
about genome function and evolution by comparing different genomes.
I focus on comparative genomics in mammals - Identification of constraint elements, regulatory motif evolution, and
lineage specific conservation. We recently analyzed data from 29 eutherian mammals, with an emphasis on annotating the human genome. The number of available mammalian
genomes is expected to grow into hundreds, and eventually thousands, within the next few years.
This yields both new oppurtunities and algorithmic and stasticial challenges due to the huge size of the datasets.
I also use data on human variation (e.g. the 1000 genomes project) to compare recent selection in the human lineage to more ancient selection in mammals.
Compressed Sensing - theory and applications in genomics.
This new theory provides means for efficient data equisition in various domains, using sparse representation of the underlying signal.
I focus on practical and theoretical problems arising from genomics applications.
One project is of group testing - identifying a small set of individuals carrying a rare allele using pooled sequencing experiments.
Another application is in metagenomics - I study the problem of reconstructing the bacterial species present in a sample.
I am also actively seeking other problems in computational biology in which sparse representation is beneficial.