Population genetics of identity by descent

Population genetics of identity by descent
Pier Francesco Palamara, Ph.D. thesis

Recent improvements in high-throughput genotyping and sequencing technologies have afforded the collection of massive, genome-wide datasets of DNA information from hundreds of thousands of individuals. These datasets, in turn, provide unprecedented opportunities to reconstruct the history of human populations and detect genotype-phenotype association. Recently developed computational methods can identify long-range chromosomal segments that are identical across samples, and have been transmitted from common ancestors that lived tens to hundreds of generations in the past. These segments reveal genealogical relationships that are typically unknown to the carrying individuals. In this work, we demonstrate that such identical-by-descent (IBD) segments are informative about a number of relevant population genetics features: they enable the inference of details about past population size fluctuations, migration events, and they carry the genomic signature of natural selection. We derive a mathematical model, based on coalescent theory, that allows for a quantitative description of IBD sharing across purportedly unrelated individuals, and develop inference procedures for the reconstruction of recent demographic events, where classical methodologies are statistically underpowered. We analyze IBD sharing in several contemporary human populations, including representative communities of the Jewish Diaspora, Kenyan Maasai samples, and individuals from several Dutch provinces, in all cases retrieving evidence of fine-scale demographic events from recent history. Finally, we expand the presented model to describe distributions for those sites in IBD shared segments that harbor mutation events, showing how these may be used for the inference of mutation rates in humans and other species.

Advertisements

Consider public archiving for your dissertation

This guest post is by Carl Boettiger (@cboettig). Carl is a postdoc with interests in theoretical and applied ecology, evolution, and phylogenetics. He’s a supporter of open access and open science, and recently posted his PhD thesis to figshare (see discussion with him on the merits of theses on figshare and University archives here).

Consider public archiving for your dissertation

As researchers we spend an immense amount of time generating products other than papers. While we go through great lengths to see that our papers are published in just the right place to be seen by our colleagues (fretting about the different impact factors, percieved audience, editorial boards, open access policies, and many other factors that determine just how a paper will see the light of day), other products of our labors largely languish on forgotten hard-drives from long ago.

Among the items that recieve considerable investiment of blood, sweat and tears in not only producing but formating just right, etc, is the PhD dissertation. As much of this work will no doubt eventually make its way into various formal publications, if it hasn’t already, it easy to view the process more as ritual than practical, whose only outcome will be another dusty black cover to grace the darkest shelves of the University library and the office of any adviser over fifty. Yet dissertations have more practical uses than bookends
as well.

A dissertation is frequently the first time certain results will see the light of day, and may offer a more accessible introduction with more complete review of background material than a published paper, thanks to the long-hand monograph style that seems to be out of vogue in the peer reviewed literature. Dissertation acknowledgements often provide wonderful snapshot into the toils of a PhD in recognizing contributions and support. And while the published results may appear only in journals requiring subscriptions, the author can almost always still release the original thesis as open access to gain the potential benefits of larger readership.[1]

While some dissertations have been important references to me during my own PhD and beyond, they aren’t always easy to find — for me, author’s webpages have been a more common source than University or publisher catalogs. Meanwhile, many other researchers do not even mention their dissertations on their own websites. Today, there are better and easier alternatives for sharing your dissertation.

An increasing recognition of other products of research has led to a proliferation of possible outlets to share research materials. Repositories such as arXiv and Figshare are indexed by Google Scholar, provide reliable persistent storage, and permanent identifiers or DOIs that can make it easy to cite or link.

[1]: e.g. see:
1. Gargouri, Y. et al. Self-Selected or Mandated, Open Access Increases Citation Impact for Higher Quality Research. PLoS ONE 5, e13636 (2010).
2. Eysenbach, G. Citation advantage of open access articles. PLoS Biology 4, e157 (2006).