Integrating Multi-Omics with environmental data for precision health: A novel analytic framework and case study on prenatal mercury induced childhood fatty liver disease

BACKGROUND

Precision Health aims to revolutionize disease prevention by leveraging information across multiple omic datasets (multi-omics). However, existing methods generally do not consider personalized environmental risk factors (e.g., environmental pollutants).

OBJECTIVE

To develop and apply a precision health framework which combines multiomic integration (including early, intermediate, and late integration, representing sequential stages at which omics layers are combined for modeling) with mediation approaches (including high-dimensional mediation to identify biomarkers, mediation with latent factors to identify pathways, and integrated/quasi-mediation to identify high-risk subpopulations) to identify novel biomarkers of prenatal mercury induced metabolic dysfunction-associated fatty liver disease (MAFLD), elucidate molecular pathways linking prenatal mercury with MAFLD in children, and identify high-risk children based on integrated exposure and multiomics data.

METHODS

This prospective cohort study used data from 420 mother-child pairs from the Human Early Life Exposome (HELIX) project. Mercury concentrations were determined in maternal or cord blood from pregnancy. Cytokeratin 18 (CK-18; a MAFLD biomarker) and five omics layers (DNA Methylation, gene transcription, microRNA, proteins, and metabolites) were measured in blood in childhood (age 6-10 years).

RESULTS

Each standard deviation increase in prenatal mercury was associated with a 0.11 [95% confidence interval: 0.02-0.21] standard deviation increase in CK-18. High dimensional mediation analysis identified 10 biomarkers linking prenatal mercury and CK-18, including six CpG sites and four transcripts. Mediation with latent factors identified molecular pathways linking mercury and MAFLD, including altered cytokine signaling and hepatic stellate cell activation. Integrated/quasi-mediation identified high risk subgroups of children based on unique combinations of exposure levels, omics profiles (driven by epigenetic markers), and MAFLD.

CONCLUSIONS

Prenatal mercury exposure is associated with elevated liver enzymes in childhood, likely through alterations in DNA methylation and gene expression. Our analytic framework can be applied across many different fields and serve as a resource to help guide future precision health investigations.

DP2LM: leveraging deep learning approach for estimation and hypothesis testing on mediation effects with high-dimensional mediators and complex confounders

Traditional linear mediation analysis has inherent limitations when it comes to handling high-dimensional mediators. Particularly, accurately estimating and rigorously inferring mediation effects is challenging, primarily due to the intertwined nature of the mediator selection issue. Despite recent developments, the existing methods are inadequate for addressing the complex relationships introduced by confounders. To tackle these challenges, we propose a novel approach called DP2LM (Deep neural network-based Penalized Partially Linear Mediation). This approach incorporates deep neural network techniques to account for nonlinear effects in confounders and utilizes the penalized partially linear model to accommodate high dimensionality. Unlike most existing works that concentrate on mediator selection, our method prioritizes estimation and inference on mediation effects. Specifically, we develop test procedures for testing the direct and indirect mediation effects. Theoretical analysis shows that the tests maintain the Type-I error rate. In simulation studies, DP2LM demonstrates its superior performance as a modeling tool for complex data, outperforming existing approaches in a wide range of settings and providing reliable estimation and inference in scenarios involving a considerable number of mediators. Further, we apply DP2LM to investigate the mediation effect of DNA methylation on cortisol stress reactivity in individuals who experienced childhood trauma, uncovering new insights through a comprehensive analysis.

Identifying pleiotropic genes via the composite test amidst the complexity of polygenic traits

Identifying the causal relationship between genotype and phenotype is essential to expanding our understanding of the gene regulatory network spanning the molecular level to perceptible traits. A pleiotropic gene can act as a central hub in the network, influencing multiple outcomes. Identifying such a gene involves testing under a composite null hypothesis where the gene is associated with, at most, one trait. Traditional methods such as meta-analyses of top-hit $P$-values and sequential testing of multiple traits have been proposed, but these methods fail to consider the background of genome-wide signals. Since Huang's composite test produces uniformly distributed $P$-values for genome-wide variants under the composite null, we propose a gene-level pleiotropy test that entails combining the aforementioned method with the aggregated Cauchy association test. A polygenic trait involves multiple genes with different functions to co-regulate mechanisms. We show that polygenicity should be considered when identifying pleiotropic genes; otherwise, the associations polygenic traits initiate will give rise to false positives. In this study, we constructed gene-trait functional modules using the results of the proposed pleiotropy tests. Our analysis suite was implemented as an R package PGCtest. We demonstrated the proposed method with an application study of the Taiwan Biobank database and identified functional modules comprising specific genes and their co-regulated traits.

Adaptive bootstrap tests for composite null hypotheses in the mediation pathway analysis

Mediation analysis aims to assess if, and how, a certain exposure influences an outcome of interest through intermediate variables. This problem has recently gained a surge of attention due to the tremendous need for such analyses in scientific fields. Testing for the mediation effect (ME) is greatly challenged by the fact that the underlying null hypothesis (i.e. the absence of MEs) is composite. Most existing mediation tests are overly conservative and thus underpowered. To overcome this significant methodological hurdle, we develop an adaptive bootstrap testing framework that can accommodate different types of composite null hypotheses in the mediation pathway analysis. Applied to the product of coefficients test and the joint significance test, our adaptive testing procedures provide type I error control under the composite null, resulting in much improved statistical power compared to existing tests. Both theoretical properties and numerical examples of the proposed methodology are discussed.

Model-based hypothesis tests for the causal mediation of semi-competing risks

Analyzing the causal mediation of semi-competing risks has become important in medical research. Semi-competing risks refers to a scenario wherein an intermediate event may be censored by a primary event but not vice versa. Causal mediation analyses decompose the effect of an exposure on the primary outcome into an indirect (mediation) effect: an effect mediated through a mediator, and a direct effect: an effect not through the mediator. Here we proposed a model-based testing procedure to examine the indirect effect of the exposure on the primary event through the intermediate event. Under the counterfactual outcome framework, we defined a causal mediation effect using counting process. To assess statistical evidence for the mediation effect, we proposed two tests: an intersection-union test (IUT) and a weighted log-rank test (WLR). The test statistic was developed from a semi-parametric estimator of the mediation effect using a Cox proportional hazards model for the primary event and a series of logistic regression models for the intermediate event. We built a connection between the IUT and WLR. Asymptotic properties of the two tests were derived, and the IUT was determined to be a size [Formula: see text] test and statistically more powerful than the WLR. In numerical simulations, both the model-based IUT and WLR can properly adjust for confounding covariates, and the Type I error rates of the proposed methods are well protected, with the IUT being more powerful than the WLR. Our methods demonstrate the strongly significant effects of hepatitis B or C on the risk of liver cancer mediated through liver cirrhosis incidence in a prospective cohort study. The proposed method is also applicable to surrogate endpoint analyses in clinical trials.

Genome-wide multimediator analyses using the generalized Berk-Jones statistics with the composite test

MOTIVATION

Mediation analysis is performed to evaluate the effects of a hypothetical causal mechanism that marks the progression from an exposure, through mediators, to an outcome. In the age of high-throughput technologies, it has become routine to assess numerous potential mechanisms at the genome or proteome scales. Alongside this, the necessity to address issues related to multiple testing has also arisen. In a sparse scenario where only a few genes or proteins are causally involved, conventional methods for assessing mediation effects lose statistical power because the composite null distribution behind this experiment cannot be attained. The power loss hence decreases the true mechanisms identified after multiple testing corrections. To fairly delineate a uniform distribution under the composite null, Huang (Genome-wide analyses of sparse mediation effects under composite null hypotheses. Ann Appl Stat 2019a;13:60-84; AoAS) proposed the composite test to provide adjusted P-values for single-mediator analyses.

RESULTS

Our contribution is to extend the method to multimediator analyses, which are commonly encountered in genomic studies and also flexible to various biological interests. Using the generalized Berk-Jones statistics with the composite test, we proposed a multivariate approach that favors dense and diverse mediation effects, a decorrelation approach that favors sparse and consistent effects, and a hybrid approach that captures the edges of both approaches. Our analysis suite has been implemented as an R package MACtest. The utility is demonstrated by analyzing the lung adenocarcinoma datasets from The Cancer Genome Atlas and Clinical Proteomic Tumor Analysis Consortium. We further investigate the genes and networks whose expression may be regulated by smoking-induced epigenetic aberrations.

AVAILABILITY AND IMPLEMENTATION

An R package MACtest is available on https://github.com/roqe/MACtest.

Exploring the relationship between socioeconomic deprivation index and Alzheimer's disease using summary-level data: From genetic correlation to causality

Patients with Alzheimer's disease (AD) are markedly increasing as population aging and no disease-modifying therapies are currently available for AD. Previous studies suggested a broad link between socioeconomic status and a variety of disorders, including mental illness and cognitive abilities. However, the association between socioeconomic deprivation and AD has been unknown. We here employed Townsend deprivation index (TDI) to explore such relation and found a positive genetic correlation (r̂_g=0.211, P = 8.00 × 10^-4) between the two traits with summary statistics data (N = 455,258 for TDI and N = 455,815 for AD). Then, we performed pleiotropy analysis at both variant and gene levels using a powerful method called PLACO and detected 87 distinct pleiotropic genes. Functional analysis demonstrated these genes were significantly enriched in pancreas, liver, heart, blood, brain, and muscle tissues. Using Mendelian randomization methods, we further found that one genetically predicted standard deviation elevation in TDI could lead to approximately 18.5% (95% confidence intervals 1.6- 38.2%, P = 0.031) increase of AD risk, and that the identified causal association was robust against used MR approaches, horizontal pleiotropy, and instrumental selection. Overall, this study provides deep insight into common genetic components underlying TDI and AD, and further reveals causal connection between them. It is also helpful to develop a more suitable plan for ameliorating inequities, hardship, and disadvantage, with the hope of improving health outcomes among economically disadvantaged people.

PhyloMed: a phylogeny-based test of mediation effect in microbiome

Microbiome data from sequencing experiments contain the relative abundance of a large number of microbial taxa with their evolutionary relationships represented by a phylogenetic tree. The compositional and high-dimensional nature of the microbiome mediator challenges the validity of standard mediation analyses. We propose a phylogeny-based mediation analysis method called PhyloMed to address this challenge. Unlike existing methods that directly identify individual mediating taxa, PhyloMed discovers mediation signals by analyzing subcompositions defined on the phylogenic tree. PhyloMed produces well-calibrated mediation test p-values and yields substantially higher discovery power than existing methods.

Genetic correlation and gene-based pleiotropy analysis for four major neurodegenerative diseases with summary statistics

Recent genome-wide association studies suggested shared genetic components between neurodegenerative diseases. However, pleiotropic association patterns among them remain poorly understood. We here analyzed 4 major neurodegenerative diseases including Alzheimer's disease (AD), Parkinson's disease (PD), frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS), and found suggestively positive genetic correlation. We next implemented a gene-centric pleiotropy analysis with a powerful method called PLACO and detected 280 pleiotropic associations (226 unique genes) with these diseases. Functional analyses demonstrated that these genes were enriched in the pancreas, liver, heart, blood, brain, and muscle tissues; and that 42 pleiotropic genes exhibited drug-gene interactions with 341 drugs. Using Mendelian randomization, we discovered that AD and PD can increase the risk of developing ALS, and that AD and ALS can also increase the risk of developing FTD, respectively. Overall, this study provides in-depth insights into shared genetic components and causal relationship among the 4 major neurodegenerative diseases, indicating genetic overlap and causality commonly drive their co-occurrence. It also has important implications on the etiology understanding, drug development and therapeutic targets for neurodegenerative diseases.

How can childhood maltreatment affect post-traumatic stress disorder in adult: Results from a composite null hypothesis perspective of mediation analysis

Background

A greatly growing body of literature has revealed the mediating role of DNA methylation in the influence path from childhood maltreatment to psychiatric disorders such as post-traumatic stress disorder (PTSD) in adult. However, the statistical method is challenging and powerful mediation analyses regarding this issue are lacking.

Methods

To study how the maltreatment in childhood alters long-lasting DNA methylation changes which further affect PTSD in adult, we here carried out a gene-based mediation analysis from a perspective of composite null hypothesis in the Grady Trauma Project (352 participants and 16,565 genes) with childhood maltreatment as exposure, multiple DNA methylation sites as mediators, and PTSD or its relevant scores as outcome. We effectively addressed the challenging issue of gene-based mediation analysis by taking its composite null hypothesis testing nature into consideration and fitting a weighted test statistic.

Results

We discovered that childhood maltreatment could substantially affected PTSD or PTSD-related scores, and that childhood maltreatment was associated with DNA methylation which further had significant roles in PTSD and these scores. Furthermore, using the proposed mediation method, we identified multiple genes within which DNA methylation sites exhibited mediating roles in the influence path from childhood maltreatment to PTSD-relevant scores in adult, with 13 for Beck Depression Inventory and 6 for modified PTSD Symptom Scale, respectively.

Conclusion

Our results have the potential to confer meaningful insights into the biological mechanism for the impact of early adverse experience on adult diseases; and our proposed mediation methods can be applied to other similar analysis settings.

Methods for large-scale single mediator hypothesis testing: Possible choices and comparisons

Mediation hypothesis testing for a large number of mediators is challenging due to the composite structure of the null hypothesis,H 0 : α β = 0 ${H}_{0}:\alpha \beta =0$ (α $\alpha $ : effect of the exposure on the mediator after adjusting for confounders;β $\beta $ : effect of the mediator on the outcome after adjusting for exposure and confounders). In this paper, we reviewed three classes of methods for large-scale one at a time mediation hypothesis testing. These methods are commonly used for continuous outcomes and continuous mediators assuming there is no exposure-mediator interaction so that the productα β $\alpha \beta $ has a causal interpretation as the indirect effect. The first class of methods ignores the impact of different structures under the composite null hypothesis, namely, (1)α = 0 , β ≠ 0 $\alpha =0,\beta \ne 0$ ; (2)α ≠ 0 , β = 0 $\alpha \ne 0,\beta =0$ ; and (3)α = β = 0 $\alpha =\beta =0$ . The second class of methods weights the reference distribution under each case of the null to form a mixture reference distribution. The third class constructs a composite test statistic using the three p values obtained under each case of the null so that the reference distribution of the composite statistic is approximatelyU ( 0 , 1 ) $U(0,1)$ . In addition to these existing methods, we developed the Sobel-comp method belonging to the second class, which uses a corrected mixture reference distribution for Sobel's test statistic. We performed extensive simulation studies to compare all six methods belonging to these three classes in terms of the false positive rates (FPRs) under the null hypothesis and the true positive rates under the alternative hypothesis. We found that the second class of methods which uses a mixture reference distribution could best maintain the FPRs at the nominal level under the null hypothesis and had the greatest true positive rates under the alternative hypothesis. We applied all methods to study the mediation mechanism of DNA methylation sites in the pathway from adult socioeconomic status to glycated hemoglobin level using data from the Multi-Ethnic Study of Atherosclerosis (MESA). We provide guidelines for choosing the optimal mediation hypothesis testing method in practice and develop an R package medScan available on the CRAN for implementing all the six methods.

Mediation analysis method review of high throughput data

High-throughput technologies have made high-dimensional settings increasingly common, providing opportunities for the development of high-dimensional mediation methods. We aimed to provide useful guidance for researchers using high-dimensional mediation analysis and ideas for biostatisticians to develop it by summarizing and discussing recent advances in high-dimensional mediation analysis. The method still faces many challenges when extended single and multiple mediation analyses to high-dimensional settings. The development of high-dimensional mediation methods attempts to address these issues, such as screening true mediators, estimating mediation effects by variable selection, reducing the mediation dimension to resolve correlations between variables, and utilizing composite null hypothesis testing to test them. Although these problems regarding high-dimensional mediation have been solved to some extent, some challenges remain. First, the correlation between mediators are rarely considered when the variables are selected for mediation. Second, downscaling without incorporating prior biological knowledge makes the results difficult to interpret. In addition, a method of sensitivity analysis for the strict sequential ignorability assumption in high-dimensional mediation analysis is still lacking. An analyst needs to consider the applicability of each method when utilizing them, while a biostatistician could consider extensions and improvements in the methodology.

Detecting associated genes for complex traits shared across East Asian and European populations under the framework of composite null hypothesis testing

Background

Detecting trans-ethnic common associated genetic loci can offer important insights into shared genetic components underlying complex diseases/traits across diverse continental populations. However, effective statistical methods for such a goal are currently lacking.

Methods

By leveraging summary statistics available from global-scale genome-wide association studies, we herein proposed a novel genetic overlap detection method called CONTO (COmposite Null hypothesis test for Trans-ethnic genetic Overlap) from the perspective of high-dimensional composite null hypothesis testing. Unlike previous studies which generally analyzed individual genetic variants, CONTO is a gene-centric method which focuses on a set of genetic variants located within a gene simultaneously and assesses their joint significance with the trait of interest. By borrowing the similar principle of joint significance test (JST), CONTO takes the maximum P value of multiple associations as the significance measurement.

Results

Compared to JST which is often overly conservative, CONTO is improved in two aspects, including the construction of three-component mixture null distribution and the adjustment of trans-ethnic genetic correlation. Consequently, CONTO corrects the conservativeness of JST with well-calibrated P values and is much more powerful validated by extensive simulation studies. We applied CONTO to discover common associated genes for 31 complex diseases/traits between the East Asian and European populations, and identified many shared trait-associated genes that had otherwise been missed by JST. We further revealed that population-common genes were generally more evolutionarily conserved than population-specific or null ones.

Conclusion

Overall, CONTO represents a powerful method for detecting common associated genes across diverse ancestral groups; our results provide important implications on the transferability of GWAS discoveries in one population to others.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-022-03637-8.

Efficient estimation of indirect effects in case-control studies using a unified likelihood framework

Mediation models are a set of statistical techniques that investigate the mechanisms that produce an observed relationship between an exposure variable and an outcome variable in order to deduce the extent to which the relationship is influenced by intermediate mediator variables. For a case-control study, the most common mediation analysis strategy employs a counterfactual framework that permits estimation of indirect and direct effects on the odds ratio scale for dichotomous outcomes, assuming either binary or continuous mediators. While this framework has become an important tool for mediation analysis, we demonstrate that we can embed this approach in a unified likelihood framework for mediation analysis in case-control studies that leverages more features of the data (in particular, the relationship between exposure and mediator) to improve efficiency of indirect effect estimates. One important feature of our likelihood approach is that it naturally incorporates cases within the exposure-mediator model to improve efficiency. Our approach does not require knowledge of disease prevalence and can model confounders and exposure-mediator interactions, and is straightforward to implement in standard statistical software. We illustrate our approach using both simulated data and real data from a case-control genetic study of lung cancer.

DNA Methylation Mediates the Association Between Individual and Neighborhood Social Disadvantage and Cardiovascular Risk Factors

Low socioeconomic status (SES) and living in a disadvantaged neighborhood are associated with poor cardiovascular health. Multiple lines of evidence have linked DNA methylation to both cardiovascular risk factors and social disadvantage indicators. However, limited research has investigated the role of DNA methylation in mediating the associations of individual- and neighborhood-level disadvantage with multiple cardiovascular risk factors in large, multi-ethnic, population-based cohorts. We examined whether disadvantage at the individual level (childhood and adult SES) and neighborhood level (summary neighborhood SES as assessed by Census data and social environment as assessed by perceptions of aesthetic quality, safety, and social cohesion) were associated with 11 cardiovascular risk factors including measures of obesity, diabetes, lipids, and hypertension in 1,154 participants from the Multi-Ethnic Study of Atherosclerosis (MESA). For significant associations, we conducted epigenome-wide mediation analysis to identify methylation sites mediating the relationship between individual/neighborhood disadvantage and cardiovascular risk factors using the JT-Comp method that assesses sparse mediation effects under a composite null hypothesis. In models adjusting for age, sex, race/ethnicity, smoking, medication use, and genetic principal components of ancestry, epigenetic mediation was detected for the associations of adult SES with body mass index (BMI), insulin, and high-density lipoprotein cholesterol (HDL-C), as well as for the association between neighborhood socioeconomic disadvantage and HDL-C at FDR q < 0.05. The 410 CpG mediators identified for the SES-BMI association were enriched for CpGs associated with gene expression (expression quantitative trait methylation loci, or eQTMs), and corresponding genes were enriched in antigen processing and presentation pathways. For cardiovascular risk factors other than BMI, most of the epigenetic mediators lost significance after controlling for BMI. However, 43 methylation sites showed evidence of mediating the neighborhood socioeconomic disadvantage and HDL-C association after BMI adjustment. The identified mediators were enriched for eQTMs, and corresponding genes were enriched in inflammatory and apoptotic pathways. Our findings support the hypothesis that DNA methylation acts as a mediator between individual- and neighborhood-level disadvantage and cardiovascular risk factors, and shed light on the potential underlying epigenetic pathways. Future studies are needed to fully elucidate the biological mechanisms that link social disadvantage to poor cardiovascular health.

Hypothesis test for causal mediation of time-to-event mediator and outcome

Hepatitis B has been a well-documented risk factor of liver cancer and mortality. To what extent hepatitis B affects mortality through increasing liver cancer incidence is of research interest and remains to be studied. We formulate the research question as a hypothesis testing problem of causal mediation where both the mediator and the outcome are time-to-event variables. The problem is closely related to semicompeting risks because time to the intermediate event may be censored by an occurrence of the outcome. We propose two hypothesis testing methods: a weighted log-rank test (WLR) and an intersection-union test (IUT). A test statistic of the WLR is constructed by adapting a nonparametric estimator of the mediation effect; however, the test may be conservative regarding its Type I Error rate. To address this, we further propose the IUT, the test statistic of which is constructed under the composite null hypothesis. Asymptotic properties of the two tests are studied, showing that the IUT is a size α test with better statistical power than the WLR. The theoretical properties are supported by extensive simulation studies under finite samples. Applying the proposed methods to the motivating hepatitis study, both WLR and IUT provided strong evidence that hepatitis B had a significant mediation effect on mortality via liver cancer incidence.

A comprehensive gene-centric pleiotropic association analysis for 14 psychiatric disorders with GWAS summary statistics

BACKGROUND

Recent genome-wide association studies (GWASs) have revealed the polygenic nature of psychiatric disorders and discovered a few of single-nucleotide polymorphisms (SNPs) associated with multiple psychiatric disorders. However, the extent and pattern of pleiotropy among distinct psychiatric disorders remain not completely clear.

METHODS

We analyzed 14 psychiatric disorders using summary statistics available from the largest GWASs by far. We first applied the cross-trait linkage disequilibrium score regression (LDSC) to estimate genetic correlation between disorders. Then, we performed a gene-based pleiotropy analysis by first aggregating a set of SNP-level associations into a single gene-level association signal using MAGMA. From a methodological perspective, we viewed the identification of pleiotropic associations across the entire genome as a high-dimensional problem of composite null hypothesis testing and utilized a novel method called PLACO for pleiotropy mapping. We ultimately implemented functional analysis for identified pleiotropic genes and used Mendelian randomization for detecting causal association between these disorders.

RESULTS

We confirmed extensive genetic correlation among psychiatric disorders, based on which these disorders can be grouped into three diverse categories. We detected a large number of pleiotropic genes including 5884 associations and 2424 unique genes and found that differentially expressed pleiotropic genes were significantly enriched in pancreas, liver, heart, and brain, and that the biological process of these genes was remarkably enriched in regulating neurodevelopment, neurogenesis, and neuron differentiation, offering substantial evidence supporting the validity of identified pleiotropic loci. We further demonstrated that among all the identified pleiotropic genes there were 342 unique ones linked with 6353 drugs with drug-gene interaction which can be classified into distinct types including inhibitor, agonist, blocker, antagonist, and modulator. We also revealed causal associations among psychiatric disorders, indicating that genetic overlap and causality commonly drove the observed co-existence of these disorders.

CONCLUSIONS

Our study is among the first large-scale effort to characterize gene-level pleiotropy among a greatly expanded set of psychiatric disorders and provides important insight into shared genetic etiology underlying these disorders. The findings would inform psychiatric nosology, identify potential neurobiological mechanisms predisposing to specific clinical presentations, and pave the way to effective drug targets for clinical treatment.

Bayesian hierarchical models for high-dimensional mediation analysis with coordinated selection of correlated mediators

We consider Bayesian high-dimensional mediation analysis to identify among a large set of correlated potential mediators the active ones that mediate the effect from an exposure variable to an outcome of interest. Correlations among mediators are commonly observed in modern data analysis; examples include the activated voxels within connected regions in brain image data, regulatory signals driven by gene networks in genome data, and correlated exposure data from the same source. When correlations are present among active mediators, mediation analysis that fails to account for such correlation can be suboptimal and may lead to a loss of power in identifying active mediators. Building upon a recent high-dimensional mediation analysis framework, we propose two Bayesian hierarchical models, one with a Gaussian mixture prior that enables correlated mediator selection and the other with a Potts mixture prior that accounts for the correlation among active mediators in mediation analysis. We develop efficient sampling algorithms for both methods. Various simulations demonstrate that our methods enable effective identification of correlated active mediators, which could be missed by using existing methods that assume prior independence among active mediators. The proposed methods are applied to the LIFECODES birth cohort and the Multi-Ethnic Study of Atherosclerosis (MESA) and identified new active mediators with important biological implications.

Bayesian Sparse Mediation Analysis with Targeted Penalization of Natural Indirect Effects

Causal mediation analysis aims to characterize an exposure's effect on an outcome and quantify the indirect effect that acts through a given mediator or a group of mediators of interest. With the increasing availability of measurements on a large number of potential mediators, like the epigenome or the microbiome, new statistical methods are needed to simultaneously accommodate high-dimensional mediators while directly target penalization of the natural indirect effect (NIE) for active mediator identification. Here, we develop two novel prior models for identification of active mediators in high-dimensional mediation analysis through penalizing NIEs in a Bayesian paradigm. Both methods specify a joint prior distribution on the exposure-mediator effect and mediator-outcome effect with either (a) a four-component Gaussian mixture prior or (b) a product threshold Gaussian prior. By jointly modeling the two parameters that contribute to the NIE, the proposed methods enable penalization on their product in a targeted way. Resultant inference can take into account the four-component composite structure underlying the NIE. We show through simulations that the proposed methods improve both selection and estimation accuracy compared to other competing methods. We applied our methods for an in-depth analysis of two ongoing epidemiologic studies: the Multi-Ethnic Study of Atherosclerosis (MESA) and the LIFECODES birth cohort. The identified active mediators in both studies reveal important biological pathways for understanding disease mechanisms.

Identifying pleiotropic genes for complex phenotypes with summary statistics from a perspective of composite null hypothesis testing

Pleiotropy has important implication on genetic connection among complex phenotypes and facilitates our understanding of disease etiology. Genome-wide association studies provide an unprecedented opportunity to detect pleiotropic associations; however, efficient pleiotropy test methods are still lacking. We here consider pleiotropy identification from a methodological perspective of high-dimensional composite null hypothesis and propose a powerful gene-based method called MAIUP. MAIUP is constructed based on the traditional intersection-union test with two sets of independent P-values as input and follows a novel idea that was originally proposed under the high-dimensional mediation analysis framework. The key improvement of MAIUP is that it takes the composite null nature of pleiotropy test into account by fitting a three-component mixture null distribution, which can ultimately generate well-calibrated P-values for effective control of family-wise error rate and false discover rate. Another attractive advantage of MAIUP is its ability to effectively address the issue of overlapping subjects commonly encountered in association studies. Simulation studies demonstrate that compared with other methods, only MAIUP can maintain correct type I error control and has higher power across a wide range of scenarios. We apply MAIUP to detect shared associated genes among 14 psychiatric disorders with summary statistics and discover many new pleiotropic genes that are otherwise not identified if failing to account for the issue of composite null hypothesis testing. Functional and enrichment analyses offer additional evidence supporting the validity of these identified pleiotropic genes associated with psychiatric disorders. Overall, MAIUP represents an efficient method for pleiotropy identification.

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

Mediation analysis investigates the intermediate mechanism through which an exposure exerts its influence on the outcome of interest. Mediation analysis is becoming increasingly popular in high-throughput genomics studies where a common goal is to identify molecular-level traits, such as gene expression or methylation, which actively mediate the genetic or environmental effects on the outcome. Mediation analysis in genomics studies is particularly challenging, however, thanks to the large number of potential mediators measured in these studies as well as the composite null nature of the mediation effect hypothesis. Indeed, while the standard univariate and multivariate mediation methods have been well-established for analyzing one or multiple mediators, they are not well-suited for genomics studies with a large number of mediators and often yield conservative p-values and limited power. Consequently, over the past few years many new high-dimensional mediation methods have been developed for analyzing the large number of potential mediators collected in high-throughput genomics studies. In this work, we present a thorough review of these important recent methodological advances in high-dimensional mediation analysis. Specifically, we describe in detail more than ten high-dimensional mediation methods, focusing on their motivations, basic modeling ideas, specific modeling assumptions, practical successes, methodological limitations, as well as future directions. We hope our review will serve as a useful guidance for statisticians and computational biologists who develop methods of high-dimensional mediation analysis as well as for analysts who apply mediation methods to high-throughput genomics studies.

Large-Scale Hypothesis Testing for Causal Mediation Effects with Applications in Genome-wide Epigenetic Studies

In genome-wide epigenetic studies, it is of great scientific interest to assess whether the effect of an exposure on a clinical outcome is mediated through DNA methylations. However, statistical inference for causal mediation effects is challenged by the fact that one needs to test a large number of composite null hypotheses across the whole epigenome. Two popular tests, the Wald-type Sobel's test and the joint significant test using the traditional null distribution are underpowered and thus can miss important scientific discoveries. In this paper, we show that the null distribution of Sobel's test is not the standard normal distribution and the null distribution of the joint significant test is not uniform under the composite null of no mediation effect, especially in finite samples and under the singular point null case that the exposure has no effect on the mediator and the mediator has no effect on the outcome. Our results explain why these two tests are underpowered, and more importantly motivate us to develop a more powerful Divide-Aggregate Composite-null Test (DACT) for the composite null hypothesis of no mediation effect by leveraging epigenome-wide data. We adopted Efron's empirical null framework for assessing statistical significance of the DACT test. We showed analytically that the proposed DACT method had improved power, and could well control type I error rate. Our extensive simulation studies showed that, in finite samples, the DACT method properly controlled the type I error rate and outperformed Sobel's test and the joint significance test for detecting mediation effects. We applied the DACT method to the US Department of Veterans Affairs Normative Aging Study, an ongoing prospective cohort study which included men who were aged 21 to 80 years at entry. We identified multiple DNA methylation CpG sites that might mediate the effect of smoking on lung function with effect sizes ranging from -0.18 to -0.79 and false discovery rate controlled at level 0.05, including the CpG sites in the genes AHRR and F2RL3. Our sensitivity analysis found small residual correlations (less than 0.01) of the error terms between the outcome and mediator regressions, suggesting that our results are robust to unmeasured confounding factors.

A powerful method for pleiotropic analysis under composite null hypothesis identifies novel shared loci between Type 2 Diabetes and Prostate Cancer

There is increasing evidence that pleiotropy, the association of multiple traits with the same genetic variants/loci, is a very common phenomenon. Cross-phenotype association tests are often used to jointly analyze multiple traits from a genome-wide association study (GWAS). The underlying methods, however, are often designed to test the global null hypothesis that there is no association of a genetic variant with any of the traits, the rejection of which does not implicate pleiotropy. In this article, we propose a new statistical approach, PLACO, for specifically detecting pleiotropic loci between two traits by considering an underlying composite null hypothesis that a variant is associated with none or only one of the traits. We propose testing the null hypothesis based on the product of the Z-statistics of the genetic variants across two studies and derive a null distribution of the test statistic in the form of a mixture distribution that allows for fractions of variants to be associated with none or only one of the traits. We borrow approaches from the statistical literature on mediation analysis that allow asymptotic approximation of the null distribution avoiding estimation of nuisance parameters related to mixture proportions and variance components. Simulation studies demonstrate that the proposed method can maintain type I error and can achieve major power gain over alternative simpler methods that are typically used for testing pleiotropy. PLACO allows correlation in summary statistics between studies that may arise due to sharing of controls between disease traits. Application of PLACO to publicly available summary data from two large case-control GWAS of Type 2 Diabetes and of Prostate Cancer implicated a number of novel shared genetic regions: 3q23 (ZBTB38), 6q25.3 (RGS17), 9p22.1 (HAUS6), 9p13.3 (UBAP2), 11p11.2 (RAPSN), 14q12 (AKAP6), 15q15 (KNL1) and 18q23 (ZNF236).

We propose a new approach PLACO that uses aggregate-level genotype-phenotype association statistics—commonly referred to as GWAS summary statistics—to identify genetic variants that influence risk of two traits or diseases. It allows correlation in summary statistics between studies that may arise due to sharing of controls between disease traits. We demonstrate that PLACO can achieve major power gain over alternative methods that are typically used. We applied PLACO to Type 2 Diabetes and Prostate Cancer summary data from two large case-control studies. Many previous studies have reported an inverse association of these two chronic diseases suggesting shared risk factors; however, shared genetic mechanisms underlying this association is poorly understood. PLACO identified a number of novel shared genetic regions that are not detected by individual trait analysis. Many of the loci implicated by PLACO increase risk for one disease while decreasing risk for the other. PLACO can similarly be used on other traits to shed light on shared genetic risk factors.

Challenges Raised by Mediation Analysis in a High-Dimension Setting

BACKGROUND

Mediation analysis is used in epidemiology to identify pathways through which exposures influence health. The advent of high-throughput (omics) technologies gives opportunities to perform mediation analysis with a high-dimension pool of covariates.

OBJECTIVE

We aimed to highlight some biostatistical issues of this expanding field of high-dimension mediation.

DISCUSSION

The mediation techniques used for a single mediator cannot be generalized in a straightforward manner to high-dimension mediation. Causal knowledge on the relation between covariates is required for mediation analysis, and it is expected to be more limited as dimension and system complexity increase. The methods developed in high dimension can be distinguished according to whether mediators are considered separately or as a whole. Methods considering each potential mediator separately do not allow efficient identification of the indirect effects when mutual influences exist among the mediators, which is expected for many biological (e.g., epigenetic) parameters. In this context, methods considering all potential mediators simultaneously, based, for example, on data reduction techniques, are more adapted to the causal inference framework. Their cost is a possible lack of ability to single out the causal mediators. Moreover, the ability of the mediators to predict the outcome can be overestimated, in particular because many machine-learning algorithms are optimized to increase predictive ability rather than their aptitude to make causal inference. Given the lack of overarching validated framework and the generally complex causal structure of high-dimension data, analysis of high-dimension mediation currently requires great caution and effort to incorporate a priori biological knowledge. https://doi.org/10.1289/EHP6240.

Variance component tests of multivariate mediation effects under composite null hypotheses

Mediation effects of multiple mediators are determined by two associations: one between an exposure and mediators ( S - M ) and the other between the mediators and an outcome conditional on the exposure ( M - Y ). The test for mediation effects is conducted under a composite null hypothesis, that is, either one of the S - M and M - Y associations is zero or both are zeros. Without accounting for the composite null, the type 1 error rate within a study containing a large number of multimediator tests may be much less than the expected. We propose a novel test to address the issue. For each mediation test j , j = 1 , … , J , we examine the S - M and M - Y associations using two separate variance component tests. Assuming a zero-mean working distribution with a common variance for the element-wise S - M (and M - Y ) associations, score tests for the variance components are constructed. We transform the test statistics into two normally distributed statistics under the null. Using a recently developed result, we conduct J hypothesis tests accounting for the composite null hypothesis by adjusting for the variances of the normally distributed statistics for the S - M and M - Y associations. Advantages of the proposed test over other methods are illustrated in simulation studies and a data application where we analyze lung cancer data from The Cancer Genome Atlas to investigate the smoking effect on gene expression through DNA methylation in 15 114 genes.