Gene Mapping in the Absence of Pedigree Information
Elizabeth Thompson, University of Washington
Classical methods for locating the genes underlying heritable traits compute model-based likelihoods using data observed on members of defined pedigree structures. However, such likelihoods are constrained by the assumed pedigree structure, and by the assumption that individuals not specified as related have independent genetic data. In reality, extended multi-generation pedigrees cannot be validated from genetic data, and more importantly the unknown co-ancestry of genome among individuals not known to be related can convey substantial information. Modern genetic data allow for the detection of this co-ancestry at specific genome locations, and it is this co-ancestry of DNA affecting a trait of interest that is key to mapping the relevant genes. Recently, numerous methods for the detection of segments of genome sharing between pairs of individuals have been developed. However, combining these inferences into realized structures of the changing genome sharing across a chromosome jointly among multiple individuals has proven challenging. I will discuss a new approach to this problem, and show the extent to which it enables classical gene mapping inference in the absence of pedigree data.