Epistasis and its implications for personal genetics

Abstract

The widespread availability of high-throughput genotyping technology has opened the door to the era of personal genetics, which brings to consumers the promise of using genetic variations to predict individual susceptibility to common diseases. Despite easy access to commercial personal genetics services, our knowledge of the genetic architecture of common diseases is still very limited and has not yet fulfilled the promise of accurately predicting most people at risk. This is partly because of the complexity of the mapping relationship between genotype and phenotype that is a consequence of epistasis (gene-gene interaction) and other phenomena such as gene-environment interaction and locus heterogeneity. Unfortunately, these aspects of genetic architecture have not been addressed in most of the genetic association studies that provide the knowledge base for interpreting large-scale genetic association results. We provide here an introductory review of how epistasis can affect human health and disease and how it can be detected in population-based studies. We provide some thoughts on the implications of epistasis for personal genetics and some recommendations for improving personal genetics in light of this complexity.

Figure 1

A Simple Biochemical Systems Model that Is Consistent with a Complex Genetic Model (A) Penetrance function showing an exclusive OR (XOR) pattern of high-risk (shaded) and low-risk (unshaded) genotype combinations for two biallelic SNPs. (B) A Petri net model of a biochemical system under the control of the two SNPs from the genetic model in (A). SNP_A controls the diameter of the “arc” or “pipe” carrying molecules of type 1, which are converted to molecules of type 2 at a constant rate governed by the first “transition” (lightning) on the left (wider pipes = larger diameter). SNP_A is pleiotropic and also controls the rate at which molecules of type 2 are converted to molecules of type 3 (wider lightning bolts = faster rate). SNP_B controls the arc or pipe carrying molecules of type 2 to the second transition, which converts them to molecules of type 3. When executed as part of a threshold model, the output of this system matches the distribution of high-risk and low-risk genotypes. This Petri net model demonstrates that a simple biochemical systems model can underlie a nonlinear genetic model.

Figure 2

Panel of Genetic Markers for Type 2 Diabetes Provided by 23andMe for Each of the Authors The profile of author A (gray bars) is associated with an overall slight decrease in risk under a multiplicative model, whereas the profile of author B (black bars) is associated with a slightly increased risk. Note that all adjusted odds ratios for individual genotypes are between 0.8 and 1.2.

Figure 3

Low-Hanging and High-Hanging Genetic Fruit Under the assumption that common diseases have a complex genetic architecture, we expect there to be few SNPs with moderate to large independent and additive main effects on disease susceptibility (i.e., low-hanging fruit). Rather, most SNPs of interest will be nestled in the branches and will only be found by embracing the complexity of the genotype-to-phenotype mapping relationship that is likely to be characterized by nonlinear gene-gene interactions and other phenomena such as gene-environment interaction and locus heterogeneity.