Capturing the Spectrum of Interaction Effects in Genetic Association Studies by Simulated Evaporative Cooling Network Analysis

Abstract

Evidence from human genetic studies of several disorders suggests that interactions between alleles at multiple genes play an important role in influencing phenotypic expression. Analytical methods for identifying Mendelian disease genes are not appropriate when applied to common multigenic diseases, because such methods investigate association with the phenotype only one genetic locus at a time. New strategies are needed that can capture the spectrum of genetic effects, from Mendelian to multifactorial epistasis. Random Forests (RF) and Relief-F are two powerful machine-learning methods that have been studied as filters for genetic case-control data due to their ability to account for the context of alleles at multiple genes when scoring the relevance of individual genetic variants to the phenotype. However, when variants interact strongly, the independence assumption of RF in the tree node-splitting criterion leads to diminished importance scores for relevant variants. Relief-F, on the other hand, was designed to detect strong interactions but is sensitive to large backgrounds of variants that are irrelevant to classification of the phenotype, which is an acute problem in genome-wide association studies. To overcome the weaknesses of these data mining approaches, we develop Evaporative Cooling (EC) feature selection, a flexible machine learning method that can integrate multiple importance scores while removing irrelevant genetic variants. To characterize detailed interactions, we construct a genetic-association interaction network (GAIN), whose edges quantify the synergy between variants with respect to the phenotype. We use simulation analysis to show that EC is able to identify a wide range of interaction effects in genetic association data. We apply the EC filter to a smallpox vaccine cohort study of single nucleotide polymorphisms (SNPs) and infer a GAIN for a collection of SNPs associated with adverse events. Our results suggest an important role for hubs in SNP disease susceptibility networks. The software is available at http://sites.google.com/site/McKinneyLab/software.

Multigenic control of measles vaccine immunity mediated by polymorphisms in measles receptor, innate pathway, and cytokine genes

“…The ability to identify both single locus variants that affect immune outcome and gene-gene interactions, or even networks/pathways of interacting loci-controlling immune phenotype are increasingly common for cancer[60–62], neurologic and autoimmune disorders[63–65], and immune responses to viral vaccines including: rubella[31], smallpox[66], and now measles. Thus we are moving from single SNP association studies to multigenic analyses and eventually towards a more comprehensive, systems-level understanding of genetic influences on vaccine-induced immune responses[67].…”

Multifactor Dimensionality Reduction as a Filter-Based Approach for Genome Wide Association Studies

“…A java implementation of MDR software is publicly available through and an R implementation is available through (Winham and Motsinger-Reif, 2010). Java implementations of EC and GAIN were used from software provided by (McKinney et al, 2009). …”

Vaccinomics, adversomics, and the immune response network theory: Individualized vaccinology in the 21st century

“…The advantage of a hierarchical network structure is the potential to identify causal relationships. In addition, the no-free-lunch theorem of data mining tells us that multiple statistical methods will need to be combined to find consensus biomarkers and signatures for immune response prediction [147,148]. …”