Complex Phenotypes and the Use of Polygenic Scores to Investigate Gene × Environment Interactions for Substance Use

The chapter suggests the need of a “second-generation” candidate gene approach to adapt to first-generation limitations and to strengthen efforts to study GE interactions. The sample sizes associated with many phenotypes and areas of study in the literature (e.g., clinical trials, neuroimaging studies) are unlikely to yield sample sizes in the area of GWA and NGS studies (i.e., 200,000-300,000 participants). However, by building upon prior limitations in the candidate gene literature, using findings from GWA and NGS studies and meta-analyses, using multiple methods of analyses (e.g., gene expression analysis; methylation analysis), and using theory and prior substantive research to guide hypothesis testing, progress can be made on G X E interactions for complex phenotypes. Several illustrative path models were provided in this chapter to provide a visual frame for how we have approached G X E interactions in the past, and how, going forward, we might proceed to investigate multiple polygenic by multiple environmental models. This level of complexity may be necessary to advance the field to address the many exciting research questions of interest, as well as the challenges that confront us as we attempt to move this knowledge from discovery to practice.

An Overview of the RELIEF Algorithm and Advancements

The aim of the RELIEF algorithm is to filter out features (e.g., genes, environmental factors) that are relevant to a trait of interest, starting from a set of that may include thousands of irrelevant features. Though widely used in many fields, its application to the study of gene-environment interaction studies has been limited thus far. We provide here an overview of this machine learning algorithm and some of its variants. Using simulated data, we then compare of the performance of RELIEF to that of logistic regression for screening for gene-environment interactions in SNP data. Even though performance degrades in larger sets of markers, RELIEF remains a competitive alternative to logistic regression, and shows clear promise as a tool for the study of gene-environment interactions. Areas for further improvements of the algorithm are then suggested.

Set-Based Gene × Environment Interaction Tests for Complex Diseases with Application to Genome-Wide Association and Sequencing Studies

This chapter presents set-based approaches that focus on identifying G X E interactions rather than set-based approaches that are based primarily on detecting G main effects (e.g., via marginal effects). The author reviews both his own research and the development of his Set Based Gene EnviRonment InterAction test (SBERIA), as well as another set-based G X E approach referred to as GESAT. GESAT extends the variance component test of the SNP-set Kernel Association Test (SKAT) to evaluate G x E effects while incorporating the main SNP effects as covariates. While both of these approaches (SBERIA and GESAT) have outperformed other benchmark methods (e.g., likelihood ratio test) and have been demonstrated to retain the appropriate Type 1 error rate, in this chapter the author conducts simulation studies to compare findings for SBERIA and GESAT approaches, and identifies associated strengths and limitations of the respective methods.

Introduction and Overview of Statistical Approaches to Gene × Environment Interactions for Complex Phenotypes

This chapter provides an introduction and overview of important issues that served as motivations for this book. For many complex phenotypes (e.g., depression, diabetes, obesity, substance use), there is substantial evidence that while genetic influences are important, so are environmental influences; moreover, there is substantial evidence from both behavior genetic studies (e.g., twin and adoptee studies) and molecular genetic studies (both human and infrahuman) that genes commonly interact with environmental factors in predicting complex phenotypes. The fields of genomics and other –omics (e.g., proteomics, metabolomics) provide exciting opportunities to advance science and foster the goals of public health and a more individualized intervention approach (e.g., precision medicine). The goals of these more individualized approaches would benefit greatly not only by advances in genomics and other –omics, but also by incorporating information both on environments and their interactions with genomic and other biological material and regulatory processes (e.g., environmental signal to biological pathway responses). Such findings would thereby offer more flexible guidance to a broader range of prevention, intervention, and treatment targets, and facilitate more tailored programs based on a fuller complement of G and E influences.

A Statistical Framework for Mediation in Environmental Epigenetics

The biological effects of genes depend upon how they are expressed in target tissues at various points in time, which is determined by their epigenetic state and in turn may be influenced by the environment.Some experimental data suggests that such influences can be transmitted across generations.In this chapter, I propose a general statistical framework for modelling how environmental and germline genetic influences on disease is mediated by epigenetics, both within the individual and across generations.The approach is illustrated on simulated data and on a study of the effect of air pollution and the ARG/NOS family of genes on childhood respiratory disease.

Statistical Strategies for Modeling Gene × Environment Interactions in Longitudinal Cohort Studies

In this chapter, Mukherjee and colleagues highlight the importance of longitudinal research designs and statistical models to investigate G x E relationships. The chapter provides a history of a range of alternative statistical models that have been used to study G X E interactions with longitudinal data, as well as a critique of their relative strengths and weaknesses. Of importance, the models reviewed incorporate information on both time invariant and time varying exposures to characterize more dynamic patterns of change across time. The authors use longitudinal data from the Normative Aging Study to illustrate some of the longitudinal models by focusing on pulse pressure, which is a risk factor for arterial stiffness. A reduced set of SNPs that have been identified in this substantive area of research were identified and the outcome measure was level of lead in the tibia bone. The authors describe their approach as a pathway orientation toward identifying how environmental exposures across time may influence changes in bone lead levels. This illustration provides clarity on issues raised in modeling G X E longitudinal data that generalize to other areas of health and to other phenotypes. It also raises critical questions about where we have been and where we need to go in the modeling of G X E interactions with longitudinal data.

Marker-Set Approaches for Assessing Gene × Environment Interactions at Gene Level

This chapter reviews the statistical methods for studying GxE effect at gene-level. Depending on how the main and interaction effects are modelled, we generally classify the current methods into either fixed or random effects approaches. The fixed effects approaches model the main and interaction effects parametrically. They are easy to interpret and computationally efficient, but are sensitive to model misspecification of the multi-variant genetic and environment effects. In contrast, the random-effects approaches model the main and interaction effects in a semi-parametric fashion, and the effects are quantified using variance components. These models are more robust to model misspecification compared to fixed effects models, though with a price of increasing computational burden caused by the use of variance components to capture the effects.

Two-Stage Procedures for the Identification of Gene × Environment and Gene × Gene Interactions in Genome-Wide Association Studies

Genome-wide association studies and next generation sequencing studies offer us an unprecedented opportunity to study the genetic etiology of diseases and other traits. Over the last few years, many replicated associations between SNPs and traits have been published. It is of particular interest to identify how genes may interact with environmental factors and other genes. In this chapter, we show that a two-stage approach, where in the first stage SNPs are screened for their potential to be involved in interactions, and interactions are then tested only among SNPs that pass the screening can greatly enhance power for detecting gene-environment and gene-gene interaction in large genetic studies compared to the tests without screening.

Cumulative Gene × Environment Influences on Neurobehavioral Development in Humans and Nonhuman Primates

In this chapter Ahmed and Kinnally provide some longitudinal examples and illustrations of how G x E influences may be studied with regard to neurobehavioral (brain) development in human and non-human primates. The chapter provides keen insight into two significant conceptual and methodological issues in the study of G X E interactions. First, is the importance of considering the findings from both human and non-human studies on genes and environment, thereby suggesting a more integrative lens for thinking about, planning, and interpreting research findings in G X E research. Second, they propose the use of multiple methods to investigate G X E interactions, including in their applications the use of both SNP-based and micro-array-based methods. With the quite massive increases in available large data sources (e.g., genomics, proteomics, metabolomics), there will be clear benefits in future research to incorporate different methods or sources of data toward identifying underlying biological mechanisms. Furthermore, the use of longitudinal research designs to study G X E interactions for time-ordered change phenomena such as neurobehavioral development provides a promising approach to identify and translate basic research findings into practice.

A Gene-Based Approach for Testing Gene × Gene and Gene × Environment Interactions

The importance of the gene × gene (G × G) and gene × environment (G × E) interaction has been widely recognized. It is statistically challenging to account for interactions in the analysis of genome-wide association data. In this chapter, we introduce a gene-based method for modeling G × G and G × E interactions under the regression framework. We evaluate the type 1 error rate and power of this new method by simulations. We apply this method to the endometrial cancer case-control dataset.