Genomic Seminar - Dr. Shai Carmi
What : How many genomic segments are shared between two individuals in a population? Theory and studies in Ashkenazi Jews
Who: Dr. Shai Carmi, Columbia University NYC
When: July 26, 2012, 11:00 AM
Where: Nano seminar room, 9th floor
In isolated populations that have experienced a recent bottleneck, most pairs of individuals share recent common ancestor in several loci across their genomes, resulting in long genomic segments that are identical-by-descent (IBD). For years, the study of IBD sharing has been limited to single markers (e.g., autozygosity coefficients). New large-scale genotyping and sequencing technologies, together with new algorithms, enable now the detection of long IBD segments extending between few hundreds kb and several Mb. The ability to detect IBD sharing in large cohorts has transformed population genetics with numerous applications such as historical reconstruction, phasing, and disease mapping.
In this talk, I will present some theoretical results on IBD sharing in population models, which we derived using coalescent theory. Specifically, we obtained analytical results for the mean and variance of the total fraction of the genome shared between random pairs of individuals, as well as approximate forms for the distribution. I will show how the results can be used for (a) demongraphic reconstruction of the population history, and (b) assessment of the success rate of imputation strategies based on IBD sharing.
Finally, I will present some results for the Ashkenazi Jewish (AJ) population. It was recently found, using dense genotyping, that a random pair of AJ individuals share as much genetic material as typical third cousins. Using the distribution of shared segment lengths, this was shown to be due to a strong bottleneck of just about 300 individuals taking place about 35 generations ago. I will present initial results from "The Ashkenzi Genome Project", where whole genomes of about 50 AJ individuals were sequenced.