PCH4GeneralPed: ANOVA Estimators of Principal Components of Heritability

In KarimOualkacha/PCH4Pedigees: Principal Components of Heritability for High Dimension Quantitative Traits and General Pedigrees

Description

Estimation of the principal components of heritability in family-based designs and an arbitrary number of phenotypes using an ANOVA approach.

Usage

PCH4GeneralPed(Ped, c1.COV, c1.traits)

Arguments

Ped	Pedigree file with the first 5 columns should be in the following order: family IDs, subject IDs, father IDs, mother IDs and sex. If there are covariates in the model, they should be introduced after the five standard columns in the pedigree file followed by the quantitative traits.
c1.COV	Integer indicates the column from which the covariates begin in the Pedigree file. If there is no covariate in the pedigree file, this parameter should be set to zero.
c1.traits	Integer indicates the column from which the traits begin in the Pedigree file.

Details

This function fits a linear mixed-effects model in the formulation described in Oualkacha et al. (2012) in the case of family-based designs and an arbitrary number of phenotypes. It controlls for the family structure using a random effect and allows covariate effects as fixed effects. PCH4GeneralPed function estimates the genetic and the environmental variance components and the principal components of heritability based on an ANOVA approach. Fixed effects are estimated using generalized linear squares estimators.

We use the ridge penalized principal components of heritability proposed by Wang et al. (2007) to handle large number of traits.

To see funcionality of the PCH4GeneralPed function, you can run the two datasets examples: data(Ped.with.COVs) and data(Ped.without.COVs).

Value

nbre.families	Integer number indicates the number of families in th pedigree file.
nbre.subjects.per.family	Vector of integers. Its length equals to the total number of families used in the pedigree file. The ntype's entries are the total number of subjects within each family.
Total.study.subjects	Total subjects without missing data used in the study.
Intercept	Vector of the GLS estimate of the model intercept.
Coeff.Covariates	Vector of the GLS estimates of the covariate effects.
Var.Polygenic	Genetic variance-covariance matrix.
Var.Env	Environmental variance-covariance matrix.
Corr.Polygenic	Genetic correlation matrix.
Corr.Env	Environmental correlation matrix.
All.PCH	Ridge penalized principal components of heritability.
Regularisation.parameter	Regularisation parameter.
heritability.PCH	Vector of heritabilities of all the ridge penalized principal components of heritability.
Trait.heritabilities	Vector of single trait heritabilities.
Genetic.Variance.Proportion.PCH	Vector with entries represent the proportion of the total genetic variation explained by each principal component of heritability. The proportion of the total genetic variation explained by a gievn PCH, is defined as the ratio of the eigenvalue of the genetic variance-covariance matrix that associated with this PCH and the sum of all the eigenvalues of the genetic variance-covariance matrix.
nbr.pch	Number of the significant principal components to be hold in the analysis.

Author(s)

Karim Oualkacha et al.

References

If you use PCH4GeneralPed package in your analysis, please cite the following work:

Oualkacha, K., Labbe, A., Ciampi, A., Roy, M.A. and Maziade, M., (2012). Principal components of heritability for high dimension quantitative traits and general pedigrees. Journal of Statistical Applications in Genetics and Molecular Biology, Volume 11. Issue 2, Article 4.

Wang Y, Fang Y, Jin M (2007). A ridge penalized principal-components approach based on heritability for high-dimensional data. Hum Hered, 64, 182-191.

Ott J, Rabinowitz D (1999). A principal-components approach based on heritability for combining phenotype information. Hum Hered, 49, 106-111.

Examples

data(Ped.with.COVs)
c1.COV = 6
c1.traits = 10
PCH4GeneralPed(Ped.with.COVs, c1.COV, c1.traits)

data(Ped.without.COVs)
c1.COV = 0
c1.traits = 6
PCH4GeneralPed(Ped.without.COVs, c1.COV, c1.traits)

KarimOualkacha/PCH4Pedigees documentation built on May 20, 2019, 8:30 a.m.