Monte Carlo likelihood calculation for identity by descent data

Abstract

Two individuals are identical by descent at a genetic locus if they share the same gene copy at that locus due to inheritance from a recent common ancestor. Identity by descent can be thought of as a continuous process along the genome that is the outcome of a highly structured hidden process. The complexity of the structure rules out direct analytic methods for calculating likelihoods in most situations, so that a Monte Carlo approach is required. This thesis presents an approach that applies to many models for the underlying genetic process of crossing-over at meiosis. The method is applied to simulated data in order to examine the amount of information contained in identity by descent data about the true model for the crossing-over process and about the true relationship between the two individuals from whom the data derive. Much of the work is done with idealized continuous identity by descent data, but an extension to the Monte Carlo method is developed that allows analysis of real data. Real data consist of identity (not necessarily by descent) of gene copies at discrete locations along the genome. The method is applied to relationship inference analysis of a real data set.