Probabilistic classification of gene-by-treatment interactions on molecular count phenotypes

TL;DR

This paper introduces a new method to classify gene-by-treatment interactions in molecular traits, improving interpretation of genetic effects under treatment conditions.

Contribution

A novel Bayesian method is introduced to classify and interpret gene-by-treatment interaction types in molecular QTL data.

Findings

The method uses Bayesian model selection to assign posterior probabilities to different G×T interaction types.

Nonlinear regression on log-scale counts better captures biological relationships between genotype and molecular traits.

The approach is validated through simulations and real datasets, showing improved interpretability of G×T interactions.

Abstract

Genetic variation can modulate response to treatment (G×T) or environmental stimuli (G×E), both of which can be highly consequential in biomedicine. An effective approach to identifying G×T signals and gaining insight into molecular mechanisms is mapping quantitative trait loci (QTL) of molecular count phenotypes, such as gene expression and chromatin accessibility, under multiple treatment conditions, which is termed response molecular QTL mapping. Although standard approaches evaluate the interaction between genetics and treatment conditions, they do not distinguish between meaningful interpretations such as whether a genetic effect is observed only in the treated condition or whether a genetic effect is observed always but accentuated in the treated condition. To address this gap, we have developed a downstream method for classifying response molecular QTLs into subclasses with meaningful genetic interpretations. Our method uses Bayesian model selection and assigns posterior probabilities to different types of G×T interactions for a given feature-SNP pair. We compare linear and nonlinear regression of log ⁡ -scale counts, noting that the latter accounts for an expected biological relationship between the genotype and the molecular count phenotype. Through simulation and application to existing datasets of molecular response QTLs, we show that our method provides an intuitive and well-powered framework to report and interpret G×T interactions. We provide a software package, ClassifyGxT [1].

Responses to treatment, such as drug, therapeutic intervention, and infection, can vary across individuals at least in part due to their genetic backgrounds. Such modulation of treatment response by genotypes, or equivalently, modulation of genotype effects by treatment is called gene-by-treatment (G×T) or gene-by-environment (G×E ) interactions. Understanding G×T or G×E interactions can potentially improve strategies for prevention and treatment of diseases, for example by selecting treatments for which a patient is most likely to respond given their genetic information, or enhanced screening for individuals most susceptible to environmental exposures. Most existing methods for G×T or GxE analysis focus on detecting interactions, not classifying them. For example, they do not distinguish between the genotype effect being present only upon treatment vs the genotype effect always being present but accentuated by treatment. Herein, we propose a statistical method to classify GxT or GxE types and supply suitable software. Our method provides an intuitive way to interpret GxT or GxE interactions.

Figures