Motivation: The main topic of research in human genetics is the identification of genes and mutations that contribute to a genetic disease. One of the factors that can influence the genetic diversity in a population is natural selection. In this work, we will compare the existing long-range haplotype methods to detect selection. Our primary objective is to obtain a clear view about their power and validity under complex demographic scenarios. Methods & Results: Literature review: A systematic review that was conducted revealed five haplotype methods with available software to identify loci that have undergone selection. The five methods are the following: LRH (Sabeti, 2002; Nature 419:832-837), iHS (Voight, 2006; PLos Biol 4:72), xp-EHH (Sabeti, 2007; Nature 449:913-8), EHHST (Zhong, 2010; Hum Genet 18:1148-59) and xp-EHHST (Zhong, 2011; Stat & Its Interface 4:51–63). All of them used simulations to test their performance with the ms (Hudson, 2002; Bioinf 18:337-8) and SelSim programs (Spencer, 2004; Bioinf 20:3673-5). Sensitivity Analysis: Ms program can consider different demographic models, without selection and SelSim provides simulations under natural selection with a simple population structure. To determine the best method, we will generate simulated data using SimuPOP. SimuPOP is a forward-in-time simulation program that can construct models with selection under complex evolutionary scenarios. We will begin with an island model, a stepping stone model incorporating an environmental gradient and more complex scenarios including a hierarchically structured population. Conclusions: We will thoroughly investigate their behavior under complex scenarios. Our study sets the basis to identify the advantages and disadvantages of each method under each modeling assumption. Here, we restrict ourselves to the comparison of the methods but an extension to a model to detect selection to N populations could be developed at a later stage.

http://www.eseb2013.com/delegates/avatsiou