Statistical tests for detecting variance effects in quantitative trait studies

Dumitrascu, Bianca; Darnell, Gregory; Ayroles, Julien; Engelhardt, Barbara E.

doi:10.1093/bioinformatics/bty565

Cited by 29 publications

(41 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The DGLM was the only method examined that can accommodates variance effects arising from both the locus and from other covariates; and the locusperm method (and genomeperm, its genomewide analog) is least reliant on parametric assumptions. We would expect other methods that allow flexible modeling of covariate effects on variance to be competitive in these regards, e.g., the recent Bayesian hierarchical model of Dumitrascu et al (2018).…”

Section: Discussionmentioning

confidence: 99%

“…A number of statistical models and methods have been developed or adapted specifically to detect vQTL. These include: Levene's test (Struchalin et al 2010) and its generalizations (Soave et al 2015;Soave and Sun 2017); the Fligner-Killeen test (Fraser and Schadt 2010); Bartlett's test (Freund et al 2013); and methods based on, or related to, the double generalized linear model (DGLM) and similar (Rönnegård and Valdar 2011;Cao et al 2014;Dumitrascu et al 2018). Tests have also been developed to detect genotype associations with arbitrary functions of the phenotype, for example higher moments, and these include a variant of the Komolgorov-Smirnov test (Aschard et al 2013) and a semiparametric exponential tilt model (Hong et al 2016).…”

Section: Introductionmentioning

confidence: 99%

“…Tests have also been developed to detect genotype associations with arbitrary functions of the phenotype, for example higher moments, and these include a variant of the Komolgorov-Smirnov test (Aschard et al 2013) and a semiparametric exponential tilt model (Hong et al 2016). Of these, the ability to accommodate BVH of known source is limited to the DGLM of Rönnegård and Valdar (2011) (as well as a very recent Bayesian counterpart, described in Dumitrascu et al 2018), which can include variance effects of arbitrary covariates as well as those belonging to the target (or foreground) QTL.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Mean-Variance QTL Mapping on a Background of Variance Heterogeneity

Valdar

2018

Preprint

View full text Add to dashboard Cite

Most QTL mapping approaches seek to identify "mean QTL", genetic loci that influence the phenotype mean, after assuming that all individuals in the mapping population have equal residual variance. Recent work has broadened the scope of QTL mapping to identify genetic loci that influence phenotype variance, termed "variance QTL", or some combination of mean and variance, which we term "mean-variance QTL".Even these approaches, however, fail to address situations where some other factor, be it an environmental factor or a distant genetic locus, influences phenotype variance. We term this situation "background variance heterogeneity" (BVH) and used simulation to explore its effects on the power and false positive rate of tests for mean QTL, variance QTL, and mean-variance QTL. Specifically, we compared traditional tests, linear regression for mean QTL and Levene's test for variance QTL, with tests more recently developed, namely Cao's tests for all three types of QTL, and tests based on the double generalized linear model (DGLM), which, unlike the other approaches, explicitly models BVH. Simulations showed that, when used in conjunction with a permutation procedure, the DGLM-based tests accurately control false positive rate and are more powerful than the other tests. We also discovered that the rank-based inverse normal transform, often used to corral unruly phenotypes, can be used to mitigate the adverse effects of BVH in some circumstances. We applied the DGLM approach, which we term "mean-variance QTL mapping", to publicly available data on a mouse backcross of CAST/Ei into M16i and, after accommodating BVH driven by father, identified a new mean QTL for bodyweight at three weeks of age. KEYWORDS QTLmapping, background

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mean-Variance QTL Mapping on a Background of Variance Heterogeneity

Valdar

2018

Preprint

View full text Add to dashboard Cite

show abstract

“…However, these methods are limited to modeling binary responses 21 or major population groups 22 . Other studies tested for genotypes associated with phenotypic variance (vQTLs) 23 , but did not model population-variance relationships [24][25][26] , which generates false-positives when population variance structure exists. We show via extensive simulations that ADGLM reliably detects phenotypic variance structure and is robust to several violations of model assumptions.…”

Section: Introductionmentioning

confidence: 99%

Existence and implications of population variance structure

Musharoff

Park

Dahl

et al. 2018

Preprint

View full text Add to dashboard Cite

Identifying the genetic and environmental factors underlying phenotypic differences between populations is fundamental to multiple research communities. To date, studies have focused on the relationship between population and phenotypic mean. Here we consider the relationship between population and phenotypic variance, i.e., "population variance structure." In addition to gene-gene and gene-environment interaction, we show that population variance structure is a direct consequence of natural selection. We develop the ancestry double generalized linear model (ADGLM), a statistical framework to jointly model population mean and variance effects.We apply ADGLM to several deeply phenotyped datasets and observe ancestry-variance associations with 12 of 44 tested traits in ~113K British individuals and 3 of 14 tested traits in ~3K Mexican, Puerto Rican, and African-American individuals. We show through extensive simulations that population variance structure can both bias and reduce the power of genetic association studies, even when principal components or linear mixed models are used. ADGLM corrects this bias and improves power relative to previous methods in both simulated and real datasets. Additionally, ADGLM identifies 17 novel genotype-variance associations across six phenotypes. * :% % + : . We did this for 2000 SNPs from a sample of 100 individuals for 1000 replicates.

show abstract

“…It has long been recognized, however, that the residual variance is itself heritable (Falconer 1965;Lynch and Walsh 1998), a phenomenon that has been described theoretically (Hill and Zhang 2004;Hill and Mulder 2010), demonstrated in inbred model organisms (Sorensen et al 2015) and crops (Yang et al 2012b), and exploited in livestock improvement efforts (Mulder et al 2008;Ibáñez-Escriche et al 2008). Correspondingly, several groups have proposed statistical methods for mapping QTL controlling the extent of this residual variance, these sometimes termed "variance QTL" (vQTL) (Paré et al 2010;Valdar 2011, 2012;Cao et al 2014;Soave and Sun 2017;Dumitrascu et al 2018). However, although detection of vQTL has started to enter the mainstream of genetic analysis (Yang et al 2012a;Hulse and Cai 2013;Ayroles et al 2015;Forsberg et al 2015;Wei et al 2016;Wang and Payseur 2017;Wei et al 2017), statistical tools for this purpose remain heterogeneous.…”

Section: Introductionmentioning

confidence: 99%

`vqtl`: An`R`package for Mean-Variance QTL Mapping

Corty

Valdar

2017

Preprint

View full text Add to dashboard Cite

Existing methods for QTL mapping in experimental crosses assume that the amount of residual variation is constant across all individuals. Many common situations violate this assumption. For example, female mice often have more variable phenotypes than males, some experimenters make more precise measurements than others, and specific genetic loci may influence the extent of variation between individuals. In all such cases, the heterogeneous variance modeling approach demonstrated here provides higher power, better protection against false positives, and allows for detection of QTL that influence phenotype variance, termed vQTL. The R package vqtl makes it easy for geneticists to apply the heterogeneous variance model, control family-wide error rate (FWER), and visualize and interpret their results. Because this package is interoperable with the popular R/qtl package and uses many of the same data structures and input patterns, it will be easy for geneticists to analyze the results of their experimental crosses with vqtl, possibly discovering new QTL. Here, we demonstrate typical usage.

show abstract

Statistical tests for detecting variance effects in quantitative trait studies

Abstract: Supplementary data are available at Bioinformatics online.

Cited by 29 publications

References 53 publications

Mean-Variance QTL Mapping on a Background of Variance Heterogeneity

Mean-Variance QTL Mapping on a Background of Variance Heterogeneity

Existence and implications of population variance structure

`vqtl`: An`R`package for Mean-Variance QTL Mapping

Contact Info

Product

Resources

About

Statistical tests for detecting variance effects in quantitative trait studies

Abstract: Supplementary data are available at Bioinformatics online.

Cited by 29 publications

References 53 publications

Mean-Variance QTL Mapping on a Background of Variance Heterogeneity

Mean-Variance QTL Mapping on a Background of Variance Heterogeneity

Existence and implications of population variance structure

vqtl: AnRpackage for Mean-Variance QTL Mapping

Contact Info

Product

Resources

About

`vqtl`: An`R`package for Mean-Variance QTL Mapping