A longstanding direction of the lab: inferring shared ancestry and recent relatedness from genomes, and using that structure to study demography and map disease — with particular depth in identity-by-descent and founder populations.
← Back to research
Two people share a stretch of DNA identical by descent (IBD) when both inherited it, unbroken by recombination, from a common ancestor — so the length and abundance of shared segments record how recently, and how often, individuals and populations share ancestry. The Pe'er lab developed foundational theory as well as practical methodology for the analysis of IBD across large cohorts.
On the theory side, we showed that the length distribution of IBD segments is a quantitative readout of fine-scale demographic history, turning observed sharing into estimates of founder events, bottlenecks, and expansions. We then characterized the statistics of IBD itself — its variance under the Wright–Fisher model, and a renewal-theory description of how sharing accrues along the genome.
On the methodology side, we introduced GERMLINE, which first made genome-wide detection of long IBD segments — and thus whole-population mapping of hidden relatedness — computationally tractable, and characterized the architecture of the long-range haplotypes it detects. Building on this foundation, IBD became a tool across many problems; highlighted applications include, for instance, association mapping (DASH), homozygosity mapping in exomes, inference of historical migration and population structure, estimation of human mutation and gene-conversion rates, HLA typing, and — in a widely noted application to forensics — re-identification of genomic data through long-range familial searches.
Founder and isolated populations are powerful settings for genetics: demographic bottlenecks enrich otherwise-rare variants and lengthen shared haplotypes, sharpening the mapping of both Mendelian and complex disease.
Much of the lab's work here centers on Ashkenazi Jewish genetics. We helped build the population's genomic infrastructure — for instance, characterizing Jewish diaspora structure and shared Middle Eastern ancestry, and, with Shai Carmi, producing a high-depth whole-genome Ashkenazi reference panel that improves interpretation of personal genomes and clarifies the population's recent history. On that foundation we mapped disease in the Ashkenazi setting, with associations spanning, for example, Crohn's disease, Parkinson's disease, and schizophrenia and bipolar disorder — including ultra-rare exonic variants implicating cadherin genes in schizophrenia.
The same principles extend beyond the Ashkenazi population. The lab also worked on the isolated population of Kosrae, Micronesia, where a small founder population and pervasive relatedness power genome-wide association and IBD-based mapping of metabolic and lipid traits.
Alongside IBD and founder-population work, the lab has contributed broadly across statistical and population genetics. Highlighted contributions include, for instance, methods for pooling samples for high-throughput resequencing, detecting SNP–SNP interactions, privacy-preserving meta-analysis, calibrating association signals against bias and the Winner's Curse, detecting cohort heterogeneity, and modeling ancestry through time.