Collaborative Research: ABI Innovation: Visualization and Statistics for Spatial Population Genomic Analysis

点击翻译按钮获取中文摘要

The University of Southern California and the University of California Davis are awarded collaborative grants to create methods and software to generate maps from genomic data, as well as to solve the related problem of placing samples of unknown origin on the map. Organisms that are more closely related to each other have more similar genomes. Often, an organism and its relatives live fairly nearby, so in principle genome similarity between many different organisms can be used to construct a map of the local geography. In this map, distances correspond to how easy it is for the organism to travel; so the map will display the landscape "as the organism sees it". Specifically, the methods developed here will construct these maps in a way to allow optimization of several different criteria, and will use randomized algorithms for pattern identification to allow efficient computation on large genomic datasets. The other, related suite of methods will use emerging tools from geostatistics to place other individuals on these maps, with rigorously quantified uncertainty. The description of geographic patterns from genomic data, and location of unknown samples, will have many applications. These tools could be used to identify the origin of a newly invading parasite, or to identify key barriers dividing the range of an endangered species. They will be of interest for inferring ancestry, and might be used to locate, for instance, the likely home of a person's grandfather to within 100km or so. They can also be used to visualize the relationships between modern populations; for instance, of humans, whose modern distribution reflects a complex history of large- and small-scale migrations; or of other species for which we often know little about their migration ecology. The simple, interactive, animated graphics developed as part of this proposal will help give researchers a new way of displaying informationand contribute to the field of next-generation data visualization. To further enable effective communication of genetic results, short workshops will be offered both locally and at national meetings, with the post-doctoral fellow and/or graduate students hired through this proposal taking a leading role. The resulting tools will be distributed and discussed at http://ralphlab.usc.edu/ and http://gcbias.org/.