trainingCovariates: Covariates of the population structure for 229 individuals

In AffiXcan: A Functional Approach To Impute Genetically Regulated Expression

Description

Toy data used in examples to describe affiXcanTrain() function.

Usage

data(trainingCovariates)

Format

An object of class data.frame

Details

This object consists in a data.frame where columns are the first three principal components of the population genetic structure and rows are individuals' IDs. These individuals are the same whom expression values are stored in the expression matrix (see help(exprMatrix) )

Genotypes of the individuals were downloaded from the GEUVADIS public dataset (https://www.ebi.ac.uk/arrayexpress/files/E-GEUV-1/) in vcf format. Following L. Price et al. (https://www.sciencedirect.com/science/article/pii/S0002929708003534), long range linkage disequilibrium (LRLD) regions were first filtered out with vcf-tools. Then, following J. Novembre et al. (www.nature.com/articles/nature07331), non-common alleles (MAF < 0.05) were filtered out with vcftools and LD pruning was performed with plink. Finally, principal components were computed with eigenstrat.

Examples

data(trainingCovariates)
head(trainingCovariates)

AffiXcan documentation built on Nov. 8, 2020, 8:07 p.m.