Gene-based mediation analysis in epigenetic studies

Abstract

<jats:title>Abstract</jats:title> <jats:p>Mediation analysis has been a useful tool for investigating the effect of mediators that lie in the path from the independent variable to the outcome. With the increasing dimensionality of mediators such as in (epi)genomics studies, high-dimensional mediation model is needed. In this work, we focus on epigenetic studies with the goal to identify important DNA methylations that act as mediators between an exposure disease outcome. Specifically, we focus on gene-based high-dimensional mediation analysis implemented with kernel principal component analysis to capture potential nonlinear mediation effect. We first review the current high-dimensional mediation models and then propose two gene-based analytical approaches: gene-based high-dimensional mediation analysis based on linearity assumption between mediators and outcome (gHMA-L) and gene-based high-dimensional mediation analysis based on nonlinearity assumption (gHMA-NL). Since the underlying true mediation relationship is unknown in practice, we further propose an omnibus test of gene-based high-dimensional mediation analysis (gHMA-O) by combing gHMA-L and gHMA-NL. Extensive simulation studies show that gHMA-L performs better under the model linear assumption and gHMA-NL does better under the model nonlinear assumption, while gHMA-O is a more powerful and robust method by combining the two. We apply the proposed methods to two datasets to investigate genes whose methylation levels act as important mediators in the relationship: (1) between alcohol consumption and epithelial ovarian cancer risk using data from the Mayo Clinic Ovarian Cancer Case-Control Study and (2) between childhood maltreatment and comorbid post-traumatic stress disorder and depression in adulthood using data from the Gray Trauma Project.</jats:p>

IUSMMT: Survival mediation analysis of gene expression with multiple DNA methylation exposures and its application to cancers of TCGA

“…We finally highlight that there is still some room for further enhancement of the power in the first stage where the relationship between a set of DNA methylation CpG sites and gene expression level was examined with the variance-component test with a linear kernel function in the present study. As well-documented in previous studies, other more complicated kernels or a composite optimal kernel may have the potential to improve the power [100][101][102][103][104][105][106]. In addition, herein we implicitly suppose that all DNA methylation CpG sites had an influence on the gene expression, which may not hold in the real-life dataset.…”

DNA methylation and aeroallergen sensitization: The chicken or the egg?

“…In a second step, HIMA estimates the mediation effect using minimax concave penalty and performs joint significance testing as a third and final step. Gene-based HMA (gHMA) We further applied gene-based high-dimensional mediation analysis (gHMA) as proposed by Fang et al [ 36 ]. The idea behind this approach is that not single CpGs but genes act as biological units and should therefore be analyzed together.…”

Statistical methods for mediation analysis in the era of high-throughput genomics: Current successes and future challenges

“…The first category of methods attempts to apply multivariate mediation methods directly on a suitable set of potential mediators or a transformed version of mediators that have a dimensionality below the sample size ( Table 1 ). This category primarily includes gene-centric methods such as gene high-dimensional mediation analysis (gHMA) [160] , principal component analysis (PCA) based methods [122] as well as several mediator-screening based methods [111] . In brief, gHMA examines one gene at a time and perform mediation analysis on the potential mediators that reside within the examined gene together.…”

“…The gHMA method adapts linear or non-linear kernels to characterize the relationship between multiple mediators and the outcome [160] . Specifically, the linear-version of gHMA (gHMA-L) examines one gene at a time and focuses on potential mediators that reside in the gene.…”

“…Afterwards, gHMA evaluates the significance of multiple mediator on outcome effects (i.e., b in Equation) through the likelihood ratio test, and further integrates the evidience of testing a and b through the same principle of JST. Besides the linear version gHMA-L, gHMA also provides an alternative version gHMA-NL that accommodates non-linear relationship between the mediators and the outcome using kernel principal components (KPC) [160] . The P values from the linear and non-linear versions of gHMA can be further combined through the ACAT (aggregated Cauchy association test) procedur to achieve robust power across various application scenarios while accounting for positive dependence among test statistics of gHMA-L and gHMA-NL [170] , [171] .…”