Abstract

Gene expression quantitative trait loci (eQTL) are genetic loci that are associated with gene expression traits. The study of the eQTL, or the genetic basis of gene expression variation, not only improves our understanding of gene expression regulation but also brings insights on the functional roles of genetic variations that influence phenotypic outcomes, such as complex human diseases. In contrast to genome-wide association studies, where the signal-to-noise ratio is often low, the eQTLs often have stronger influence on gene expression variation, and hundreds or thousands of eQTLs may be recovered. We conjecture that one of the major applications of eQTL findings is to construct directed graphical models of gene expression data. In this chapter, we review the methods for eQTL mapping, constructing directed graphical models, and the approaches to construct directed graphical models using eQTL data.

Introduction

The expression of a gene may be associated with the genotype of one or more genetic loci, and such loci are often referred to as gene expression quantitative trait loci (eQTLs). An eQTL study is an integrated study of genetic variants and gene expression across a group of samples. In many eQTL studies, phenotype data (e.g., disease status or drug response) are also collected, and it is of great interest to use eQTLresults to inform or guide the phenotype study.Apromising approach toward this goal is to construct a directed gene-gene network using eQTL data. In this chapter, we provide reviews and discussions on constructing directed graphical models using eQTL data.

It has been well appreciated that a gene network perspective is crucial to understanding the molecular basis of complex traits, such as many human diseases (Barabási et al., 2011; Marbach et al., 2012). Gene networks can be studied by undirected or directed graphs. For example, a protein-protein interaction graph, where two proteins are connected if they interact with each other, is an undirected graph. A biological pathway often corresponds to a directed graph. The meaning of a directed edge within a pathway depends on the nature of the pathway. In a gene regulation pathway, an edge A → B indicates A regulates B. In a signaling pathway, an edge A → B indicates signal is transmitted from A to B. Pathway-level analysis is a crucial step to understanding the molecular basis of complex traits, including many human diseases.