Analysis of Genetic, Parent of Origin or Treatment effect on gene expression using RNA-seq data in Human and Mouse

RNA sequencing allows us to systematically study allelic imbalance of gene expression, which may be due to genetic factors or genomic imprinting. In order to avoid confounding between genetic and parent-of-origin effects, and to improve the power to detect either effect, we have developed new statistical methods to jointly model both effects. In this dissertation, we consider a situation where modeling and separation of genetic and parent-of-origin effects are more challenging. First, we consider outbred populations such as human. We propose to collect RNA-seq data from children of family trios as well as phased genotype data for each member of those trios. Then we capture the genetic effects by cis-acting eQTLs and use the phased genotype data to define parent-of-origin effects. Next we propose a protocol for processing and analysis of RNAseq data with proper integration of total and allele-specific counts. We compare two major methods for final analysis as well as propose an efficient method for estimating permutation p-value. Finally we study for treatment, sex and additive genetic effect the reciprocal inbred crosses (RIX) produced from eight divergent inbred strains.Doctor of Public Healt