Multi-SNP Mediation Intersection-Union Test

Zhong, Wujuan; Spracklen, Cassandra N.; Mohlke, Karen L.; Zheng, Xiaojing; Fine, Jason P.; Li, Yun

doi:10.1101/455352

Cited by 6 publications

(20 citation statements)

References 51 publications

(65 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Huang et al's methods are kernel-based regression methods and use variance component score statistic to test for mediation but these methods assume a priori known expression quantitative trait loci (eQTLs) (Huang et al, 2015. To address lack of knowledge regarding eQTLs, we have extended Baron and Kenny's framework to handle mediation effect of high-dimensional genetic variants on a continuous outcome (Zhong et al, 2019). To the best of our knowledge, none of the existing methods can jointly test mediation effects of multiple correlated SNPs on a non-Gaussian outcome.…”

Section: Introductionmentioning

confidence: 99%

“…We propose a generalized multi-SNP mediation intersection-union test to accommodate both mediation and direct effects of multiple correlated SNPs on non-Gaussian outcomes without a prior knowledge of eQTLs. Similar to our previously developed SMUT method (Zhong et al, 2019), the method proposed in this work is an extension of Baron and Kenny's framework and leverages intersection-union test (IUT) to decompose mediation into two separate regression models. Our proposed method SMUT GLM and SMUT PH deals with two categories of non-Gaussian outcomes.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Generalized Multi-SNP Mediation Intersection-Union Test

Zhong

Darville

Zheng

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

To elucidate the molecular mechanisms underlying genetic variants identified from genome-wide association studies (GWAS) for a variety of phenotypic traits encompassing binary, continuous, count, and survival outcomes, we propose a novel and flexible method to test for mediation that can simultaneously accommodate multiple genetic variants and different types of outcome variables. Specifically, we employ the intersection-union test approach combined with likelihood ratio test to detect mediation effect of multiple genetic variants via some mediator (for example, the expression of a neighboring gene) on outcome. We fit high-dimensional generalized linear mixed models under the mediation framework, separately under the null and alternative hypothesis. We leverage Laplace approximation to compute the marginal likelihood of outcome and use coordinate descent algorithm to estimate corresponding parameters. Our extensive simulations demonstrate the validity of our proposed method and substantial, up to 97%, power gains over alternative methods. Applications to real data for the study of Chlamydia trachomatis infection further showcase advantages of our method. We believe our proposed method will be of value and general interest in this post-GWAS era to disentangle the potential causal mechanism from DNA to phenotype for new drug discovery and personalized medicine.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Generalized Multi-SNP Mediation Intersection-Union Test

Zhong

Darville

Zheng

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…If MR were naively applied for each of these investigate the properties and performance of the different approaches. We also 143 applied two recently-proposed methods, LCV [26] and SMUT [41], along with BN, MR 144 and a recent MR extension [22], to data generated under a more complex simulation 145 scenario involving extreme pleiotropy. 146 Illustrative Example: fatty acids and BMI 147 As an initial motivating example, we investigated possible causal relationships between 148 fatty acid metabolites and body mass index (BMI) using the TwinsUK study data [42].…”

mentioning

confidence: 99%

“…Arguably the most 591 serious limitation of BN is the fact that analysis is performed on individual level data, 592 and the method is not readily extended to summary data (although this represents an 593 interesting topic for future investigation). In contrast, MR approaches such as 594 Multi-SNP Mediation Intersection-Union Test 706 We also applied a recently-proposed multi-SNP mediation intersection-union test 707 known as SMUT [41]. SMUT tests the joint mediation effects of multiple (potentially 708 correlated) genetic variants on a trait through a single mediator, effectively generating 709 a hypothesis test for mediation but with a multivariate exposure.…”

mentioning

confidence: 99%

“…Weighted allele 836 score variables were used as instrumental variables and were re-constructed within 837 each simulation replicate using SNPs passing a p-value threshold of p < 5 × 10 −6 of 838 association with the appropriate metabolite or with Y, using the estimated regression 839 coefficients as weights. The same SNPs were also used as the genetic variants in 840 SMUT (using the R package SMUT [41]), which was also performed between every 841 individual metabolite and Y (or vice versa for reverse causal effects).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Bayesian network analysis incorporating genetic anchors complements conventional Mendelian randomization approaches for exploratory analysis of causal relationships in complex data

Howey

Shin

Relton

et al. 2019

Preprint

View full text Add to dashboard Cite

Mendelian randomization (MR) is an increasingly popular causal inference tool used in genetic epidemiology. But it can have limitations for evaluating simultaneous causal relationships in complex data sets that include, for example, multiple genetic predictors and multiple potential risk factors associated with the same genetic variant. Here we use real and simulated data to investigate Bayesian network analysis (BN) as an alternative approach. A Bayesian network describes the conditional dependencies/independencies of variables using a graphical model (a directed acyclic graph) and its accompanying joint probability. In real data, we found BN inferred simultaneous causal relationships that confirmed the individual causal relationships suggested by bi-directional MR, while allowing for the existence of potential horizontal pleiotropy (that would violate MR assumptions). In simulated data, BN with two directional anchors (mimicking genetic instruments) had greater power for a fixed type 1 error than bi-directional MR, while BN with a single directional anchor performed better than or as well as bi-directional MR. Both BN and MR could be adversely affected by violations of their underlying assumptions (such as genetic confounding due to unmeasured horizontal pleiotropy). BN with no directional anchor generated inference that was no better than by chance, emphasizing the importance of directional anchors in BN (as in MR). Under highly pleiotropic simulated scenarios, BN outperformed both MR (and its recent extensions) and two recently-proposed alternative approaches: a multi-SNP mediation intersection-union test (SMUT) and a latent causal variable (LCV) test. We conclude that BN is a useful complementary method to MR for performing causal inference in complex data sets such as those generated from modern "omics" technologies Author summaryMendelian randomization (MR) is a popular method for inferring causal relationships between variables (such as between an intermediate biological factor and a disease outcome). However, MR relies on a number of assumptions that may be hard to verify, and it is not ideally suited to comparing different underlying causal scenarios. Here we September 19, 2019 1/52 propose the use of an alternative method, Bayesian network analysis (BN), as a complementary tool to MR. We use real and simulated data to investigate the performance of MR, BN and several other recently-proposed methods, and find that BN performs as well as, or better than, the other methods, particularly under complex scenarios. We conclude that BN is a useful complementary method to MR for performing causal inference in complex data sets. 35 been applied to identify innovative drug targets in early stage drug development or to 36 discover novel risk factors at the molecular level by scanning "omics" data 37 systemically [8-10]. An advantage of MR is that individual-level data are not 38 necessarily required; inference can be performed on the basis of summary statistics 39 measuring the relationship between the genetic ins...

show abstract

MRLocus: identifying causal genes mediating a trait through Bayesian estimation of allelic heterogeneity

Zhu

Matoba

Wilson

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Expression quantitative trait loci (eQTL) studies are used to understand the regulatory function of non-coding genome-wide association study (GWAS) risk loci, but colocalization alone does not demonstrate a causal relationship of gene expression affecting a trait. Evidence for mediation, that perturbation of gene expression in a given tissue or developmental context will induce a change in the downstream GWAS trait, can be provided by two-sample Mendelian Randomization (MR). Here, we introduce a new statistical method, MRLocus, for Bayesian estimation of the gene-to-trait effect from eQTL and GWAS summary data for loci displaying allelic heterogeneity, that is, containing multiple LD-independent eQTLs. MRLocus makes use of a colocalization step applied to each eQTL, followed by an MR analysis step across eQTLs. Additionally, our method involves estimation of allelic heterogeneity through a dispersion parameter, indicating variable mediation effects from each individual eQTL on the downstream trait. Our method is evaluated against state-of-the-art methods for estimation of the gene-to-trait mediation effect, using an existing simulation framework. In simulation, MRLocus often has the highest accuracy among competing methods, and in each case provides more accurate estimation of uncertainty as assessed through interval coverage. MRLocus is then applied to five causal candidate genes for mediation of particular GWAS traits, where gene-to-trait effects are concordant with those previously reported. We find that MRLocus' estimation of the causal effect across eQTLs within a locus provides useful information for determining how perturbation of gene expression or individual regulatory elements will affect downstream traits. The MRLocus method is implemented as an R package available at https://mikelove.github.io/mrlocus.

show abstract

Multi-SNP Mediation Intersection-Union Test

Cited by 6 publications

References 51 publications

Generalized Multi-SNP Mediation Intersection-Union Test

Generalized Multi-SNP Mediation Intersection-Union Test

Bayesian network analysis incorporating genetic anchors complements conventional Mendelian randomization approaches for exploratory analysis of causal relationships in complex data

MRLocus: identifying causal genes mediating a trait through Bayesian estimation of allelic heterogeneity

Contact Info

Product

Resources

About