An efficient genome-wide association test for multivariate phenotypes based on the Fisher combination function

Yang, James; Li, Jia; Williams, L. Keoki; Buu, Anne

doi:10.1186/s12859-015-0868-6

Cited by 48 publications

(91 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For complex diseases with multivariate phenotypes, combining multiple related phenotypic 250 traits into one analysis can increase power in finding disease associations 75,76 . We combined 251 the six aforementioned quantitative trait associations into one meta-statistic using a Fisher 252 combination function modified to account for correlated traits 77 . Inter-trait correlations were 253 determined using the Pearson correlation coefficient ( Fig.…”

mentioning

confidence: 99%

Family-based quantitative trait meta-analysis implicates rare noncoding variants inDENND1Ain pathogenesis of polycystic ovary syndrome

Dapas¹,

Sisk²,

et al. 2018

Preprint

View full text Add to dashboard Cite

Family-based quantitative trait meta-analysis implicates rare noncoding 1 variants in DENND1A in pathogenesis of polycystic ovary syndrome 2 3 4 ABSTRACT 33Polycystic ovary syndrome (PCOS) is among the most common endocrine disorders of 34 premenopausal women, affecting 5-15% of this population depending on the diagnostic 35 criteria applied. It is characterized by hyperandrogenism, ovulatory dysfunction and 36 polycystic ovarian morphology. PCOS is a leading risk factor for type 2 diabetes in young 37 women. PCOS is highly heritable, but only a small proportion of this heritability can be 38 accounted for by the common genetic susceptibility variants identified to date. To test the 39 hypothesis that rare genetic variants contribute to PCOS pathogenesis, we performed 40 whole-genome sequencing on DNA from 62 families with one or more daughters with PCOS. 41We tested for associations of rare variants with PCOS and its concomitant hormonal traits 42 using a quantitative trait meta-analysis. We found rare variants in DENND1A (P=5.31×10 -5 , 43Padj=0.019) that were significantly associated with reproductive and metabolic traits in 44 PCOS families. Common variants in DENND1A have previously been associated with PCOS 45 diagnosis in genome-wide association studies. Subsequent studies indicated that DENND1A 46 is an important regulator of human ovarian androgen biosynthesis. Our findings provide 47 additional evidence that DENND1A plays a central role in PCOS and suggest that rare noncoding 48 variants contribute to disease pathogenesis. 50Polycystic ovary syndrome (PCOS) is a common endocrine disorder affecting 5-15% of 52 premenopausal women worldwide 1 , depending on the diagnostic criteria applied. PCOS is 53 diagnosed by two or more of its reproductive features of hyperandrogenism, ovulatory 54 dysfunction, and polycystic ovarian morphology. It is frequently associated with insulin 55 resistance and pancreatic β-cell dysfunction, making it a leading risk factor for type 2 diabetes 56 in young women 2 . 57PCOS is a highly heritable complex genetic disorder. Analogous to other complex traits 3 , 58 common susceptibility loci identified in genome-wide association studies (GWAS) 4-8 account 59 for only a small proportion of the estimated genetic heritability of PCOS 9 . As GWAS were 60 designed to assess common allelic variants, usually with minor allele frequencies (MAF) ≥2-61 5%, it has been proposed that less frequently occurring variants with greater effect sizes 2 account for the observed deficit in heritability 10 . Next generation sequencing approaches have 63 identified rare variants that contribute to complex disease pathogenesis [11][12][13][14][15][16] . 64We tested the hypothesis that rare variants contribute to PCOS by conducting family-based 65 association analyses using whole-genome sequencing data. We filtered and weighted rare 66 variants (MAF ≤2%) according to their predicted levels of deleteriousness and grouped them 67 regionally and by genes. We not only tested for associations with PCOS diagnosi...

show abstract

mentioning

confidence: 99%

Family-based quantitative trait meta-analysis implicates rare noncoding variants inDENND1Ain pathogenesis of polycystic ovary syndrome

Dapas¹,

Sisk²,

et al. 2018

Preprint

View full text Add to dashboard Cite

show abstract

“…Brown () and J. J. Yang () have shown that if these p values are correlated, the distribution of the T statistic approximates a gamma distribution with the shape parameter

v / 2

and the scale parameter

2 γ

under the null hypothesis. J. J. Yang, Li, Williams, and Buu () further extended the

T

statistic to a two‐sided test for testing pleiotropic effects. For the proposed GEE‐joint test, the

T

statistic follows a gamma distribution with a mean of 4 and a variance of

4 + δ

, where

δ = cov (- 2 l og (p_{marginal}), - 2 \log true(p_{interaction} true))

is a function of the correlation between

γ_{1}

and

β_{3}

−

ρ

.…”

Section: Methodsmentioning

confidence: 99%

“…Therefore, it can be approximated by a tenth‐order polynomial of

ρ

ρ

is estimated by the bias‐corrected sample correlation between

{\hat{γ}}_{1}

and

{\hat{β}}_{3}

(J. J. Yang et al, ).…”

Section: Methodsmentioning

confidence: 99%

Extended methods for gene–environment‐wide interaction scans in studies of admixed individuals with varying degrees of relationships

Chen

Adrianto

Ianuzzi

et al. 2019

Genetic Epidemiology

Self Cite

View full text Add to dashboard Cite

The etiology of many complex diseases involves both environmental exposures and inherited genetic predisposition as well as interactions between them. Gene–environment‐wide interaction studies (GEWIS) provide a means to identify the interactions between genetic variation and environmental exposures that underlie disease risk. However, current GEWIS methods lack the capability to adjust for the potentially complex correlations in studies with varying degrees of relationships (both known and unknown) among individuals in admixed populations. We developed novel generalized estimating equation (GEE) based methods—GEE‐adaptive and GEE‐joint—to account for phenotypic correlations due to kinship while accounting for covariates, including, measures of genome‐wide ancestry. In simulation studies of admixed individuals, both methods controlled family‐wise error rates, an advantage over the case‐only approach. They demonstrated higher power than traditional case–control methods across a wide range of underlying alternative hypotheses, especially where both marginal and interaction effects were present. We applied the proposed method to conduct a GEWIS of a known sarcoidosis risk factor (insecticide exposure) and risk of sarcoidosis in African Americans and identified two novel loci with suggestive evidence of G × E interaction.

show abstract

“…Methods involving the first strategy combine either the univariate test statistics (Kim, Bai, & Pan, 2015;O'Brien, 1984;Wei & Johnson, 1985) or p-values (Liang, Wang, Sha, & Zhang, 2016;van der Sluis, Posthuma, & Dolan, 2013;Yang, Li, Williams, & Buu, 2016); they are generally very easy to implement and can cope with a mixture of different types of phenotypes; however, the statistical power of those methods might heavily rely on the homogeneity of univariate test statistics (Zhu, Zhang, & Sha, 2015;Zhu, Zhang, & Sha, 2018); the most popular methods in this category include O'Brien's method (O'Brien, 1984;Wei & Johnson, 1985), trait-based association test that uses extended Simes procedure (TATES; van der Sluis et al, 2013), Fisher's combination (Yang et al, 2016), and adaptive Fisher's combination (AFC; Liang et al, 2016). The most widely used strategies for those research efforts are combining univariate analysis results, dimension reduction, and regression models.…”

Section: Introductionmentioning

confidence: 99%

“…The most widely used strategies for those research efforts are combining univariate analysis results, dimension reduction, and regression models. Methods involving the first strategy combine either the univariate test statistics (Kim, Bai, & Pan, 2015;O'Brien, 1984;Wei & Johnson, 1985) or p-values (Liang, Wang, Sha, & Zhang, 2016;van der Sluis, Posthuma, & Dolan, 2013;Yang, Li, Williams, & Buu, 2016); they are generally very easy to implement and can cope with a mixture of different types of phenotypes; however, the statistical power of those methods might heavily rely on the homogeneity of univariate test statistics (Zhu, Zhang, & Sha, 2015;Zhu, Zhang, & Sha, 2018); the most popular methods in this category include O'Brien's method (O'Brien, 1984;Wei & Johnson, 1985), trait-based association test that uses extended Simes procedure (TATES; van der Sluis et al, 2013), Fisher's combination (Yang et al, 2016), and adaptive Fisher's combination (AFC; Liang et al, 2016). For the strategy of dimension reduction, instead of testing one phenotype at a time, one first constructs a small number of latent variables, which are linear combinations of the observed phenotypes, and then tests the associations between the latent variables and the genetic variant of interest; dimension reduction methods are, in general, suitable only when all phenotypes are normally distributed (Yang & Wang, 2012); in addition, the newly derived latent variables are usually difficult to interpret in the real-world applications; the most popular methods in this category include principal components of the phenotypes (Aschard et al, 2014), principal component of heritability (Klei, Luca, Devlin, & Roeder, 2008;Zhou et al, 2015), and canonical correlation analysis (Ferreira & Purcell, 2008;Tang & Ferreira, 2012).…”

Section: Introductionmentioning

confidence: 99%

Joint analysis of multiple phenotypes using a clustering linear combination method based on hierarchical clustering

Zhang

Sha

2019

Genetic Epidemiology

View full text Add to dashboard Cite

Emerging evidence suggests that a genetic variant can affect multiple phenotypes, especially in complex human diseases. Therefore, joint analysis of multiple phenotypes may offer new insights into disease etiology. Recently, many statistical methods have been developed for joint analysis of multiple phenotypes, including the clustering linear combination (CLC) method. Due to the unknown number of clusters for a given data, a simulation procedure must be used to evaluate the p‐value of the final test statistic of CLC. This makes the CLC method computationally demanding. In this paper, we use a stopping criterion to determine the number of clusters in the CLC method. We have named our method, hierarchical clustering CLC (HCLC). HCLC has an asymptotic distribution, which is very computationally efficient and makes it applicable for genome‐wide association studies. Extensive simulations together with the COPDGene data analysis have been used to assess the type I error rates and power of our proposed method. Our simulation results demonstrate that the type I error rates of HCLC are effectively controlled in different realistic settings. HCLC either outperforms all other methods or has statistical power that is very close to the most powerful method with which it has been compared.

show abstract

An efficient genome-wide association test for multivariate phenotypes based on the Fisher combination function

Cited by 48 publications

References 33 publications

Family-based quantitative trait meta-analysis implicates rare noncoding variants inDENND1Ain pathogenesis of polycystic ovary syndrome

Family-based quantitative trait meta-analysis implicates rare noncoding variants inDENND1Ain pathogenesis of polycystic ovary syndrome

Extended methods for gene–environment‐wide interaction scans in studies of admixed individuals with varying degrees of relationships

Joint analysis of multiple phenotypes using a clustering linear combination method based on hierarchical clustering

Contact Info

Product

Resources

About