An efficient Bayesian meta-analysis approach for studying cross-phenotype genetic associations

Majumdar, Årindam; Haldar, Tanushree; Bhattacharya, Sourabh; Witte, John S.

doi:10.1371/journal.pgen.1007139

Cited by 48 publications

(52 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Other methods have been developed to that end. For instance, multivariate twin studies are appropriate for investigating trait etiology, or multi-trait GWAS meta-analysis aims to disentangle effects of correlated traits at the level of genetic variants (16,17,42–46). In contrast, the declared aim of the MPS approach is to maximize trait prediction, without assumptions about the relationships among predictors.…”

Section: Discussionmentioning

confidence: 99%

Multi-polygenic score approach to trait prediction

et al. 2017

View full text Add to dashboard Cite

A primary goal of polygenic scores, which aggregate the effects of thousands of trait-associated DNA variants discovered in genome-wide association studies (GWAS), is to estimate individual-specific genetic propensities and predict outcomes. This is typically achieved using a single polygenic score, but here we use a multi-polygenic score (MPS) approach to increase predictive power by exploiting the joint power of multiple discovery GWAS, without assumptions about the relationships among predictors. We used summary statistics of 81 well-powered GWAS of cognitive, medical, and anthropometric traits to predict three core developmental outcomes in our independent target sample: educational achievement, body-mass index (BMI), and general cognitive ability. We used regularized regression with repeated cross-validation to select from and estimate contributions of 81 polygenic scores in a UK-representative sample of 6,710 unrelated adolescents. The MPS approach predicted 11% variance in educational achievement, 4.8% in general cognitive ability, and 5.4% in BMI in an independent test set, improving prediction by 1.1%, 1.1%, and 1.6% variance explained over the best single-score predictions. As other relevant GWA analyses are reported, they can be incorporated in MPS models to maximize phenotype prediction. The MPS approach should be useful in research with modest sample sizes to investigate developmental, multivariate and gene-environment interplay issues and, eventually, in clinical settings to predict and prevent problems using personalized interventions.

show abstract

Section: Discussionmentioning

confidence: 99%

Multi-polygenic score approach to trait prediction

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Another advantage of our approach is that it allows one to incorporate prior or external information on the likelihood that a phenotype exhibits associations with a region via the mixing proportion, which can improve identification of associated outcomes. Our framework also extends the recently proposed CPBayes method for testing the association between a single SNP and multiple phenotypes (Majumdar et al, 2018). CPBayes imposes a spike and slab prior on the genetic SNP effect and uses a mixture of two normal distributions to represent the SNP effect under the null and alternative hypotheses.…”

Section: Discussionmentioning

confidence: 98%

“…When a single SNP is analyzed, our mixture set‐up corresponds to that of CPBayes. However, we additionally estimate the amount of heterogeneity between outcome specific associations, captured by the parameter

τ

, directly from the data, whereas in Majumdar et al () it is prespecified. Mis‐specifying the amount of heterogeneity will lower power, sensitivity, and specificity of the procedure in Majumdar et al ().…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Subset testing and analysis of multiple phenotypes

Derkach

Pfeiffer

2019

Genetic Epidemiology

View full text Add to dashboard Cite

Meta‐analysis of multiple genome‐wide association studies (GWAS) is effective for detecting single‐ or multimarker associations with complex traits. We develop a flexible procedure (subset testing and analysis of multiple phenotypes [STAMP]) based on mixture models to perform a region‐based meta‐analysis of different phenotypes using data from different GWAS and identify subsets of associated phenotypes. Our model framework helps distinguish true associations from between‐study heterogeneity. As a measure of association, we compute for each phenotype the posterior probability that the genetic region under investigation is truly associated. Extensive simulations show that STAMP is more powerful than standard approaches for meta‐analyses when the proportion of truly associated outcomes is between 25% and 50%. For other settings, the power of STAMP is similar to that of existing methods. We illustrate our method on two examples, the association of a region on chromosome 9p21 with the risk of 14 cancers, and the associations of expression of quantitative trait loci from two genetic regions with their cis‐single‐nucleotide polymorphisms measured in 17 tissue types using data from The Cancer Genome Atlas.

show abstract

“…It seems logical to expect that meta‐analyzing multiple phenotypes can further increase power of rare‐variant tests. Various methods have been developed for meta‐analysis of multiple phenotypes (Majumdar, Haldar, Bhattacharya, & Witte, ; Ray & Boehnke, ; Zhu et al, ), but most of them are single variant‐based methods, which have low power to identify rare‐variant associations. More powerful gene or region‐based tests for multiple phenotypes have been developed for use within a single study (Broadaway et al, ; Selyeong Lee et al, ; B. Wu & Pankow, ).…”

Section: Introductionmentioning

confidence: 99%

Meta‐MultiSKAT: Multiple phenotype meta‐analysis for region‐based association test

Dutta

Taliun

Weinstock

et al. 2019

Genetic Epidemiology

View full text Add to dashboard Cite

The power of genetic association analyses can be increased by jointly meta‐analyzing multiple correlated phenotypes. Here, we develop a meta‐analysis framework, Meta‐MultiSKAT, that uses summary statistics to test for association between multiple continuous phenotypes and variants in a region of interest. Our approach models the heterogeneity of effects between studies through a kernel matrix and performs a variance component test for association. Using a genotype kernel, our approach can test for rare‐variants and the combined effects of both common and rare‐variants. To achieve robust power, within Meta‐MultiSKAT, we developed fast and accurate omnibus tests combining different models of genetic effects, functional genomic annotations, multiple correlated phenotypes, and heterogeneity across studies. In addition, Meta‐MultiSKAT accommodates situations where studies do not share exactly the same set of phenotypes or have differing correlation patterns among the phenotypes. Simulation studies confirm that Meta‐MultiSKAT can maintain the type‐I error rate at the exome‐wide level of 2.5 × 10−6. Further simulations under different models of association show that Meta‐MultiSKAT can improve the power of detection from 23% to 38% on average over single phenotype‐based meta‐analysis approaches. We demonstrate the utility and improved power of Meta‐MultiSKAT in the meta‐analyses of four white blood cell subtype traits from the Michigan Genomics Initiative (MGI) and SardiNIA studies.

show abstract

An efficient Bayesian meta-analysis approach for studying cross-phenotype genetic associations

Cited by 48 publications

References 42 publications

Multi-polygenic score approach to trait prediction

Multi-polygenic score approach to trait prediction

Subset testing and analysis of multiple phenotypes

Meta‐MultiSKAT: Multiple phenotype meta‐analysis for region‐based association test

Contact Info

Product

Resources

About