Bayesian linkage analysis for autotetraploids

Eunjung Lim, Purdue University

Abstract

Polyploids are species with more than two copies of each chromosome in their genome. These additional copies make constructing a genetic linkage map far more challenging than for diploids because there is incomplete parental configuration information, incomplete progeny genotype information, and a more complicated meiosis process. These complexities and limited information create additional uncertainty in the genetic map. In practice, however, a single linkage map is commonly estimated and then treated as fixed in further analyses, such as QTL analysis. For polyploids, we show that using a single map can lead to incorrect results or results stated with too much confidence. Thus, we develop a Bayesian approach to linkage mapping so these additional uncertainties can be included in the posterior distribution of the genetic map. We develop a Markov chain Monte Carlo (MCMC) algorithm to account for the unknown parental configuration, order of markers along a chromosome, and their recombination fractions. Given the posterior distribution of the genetic linkage map, we then propose using this distribution when performing QTL interval mapping so that map uncertainty is properly addressed. The performance of this proposed method is assessed by simulation study and analysis of a real alfalfa data set. The results are also compared to TetraploidMap, a free program that builds a genetic linkage map for autotetraploids and performs interval QTL mapping.

Degree

Ph.D.

Advisors

Craig, Purdue University.

Subject Area

Biostatistics|Genetics

COinS