R/data.R

In julstpierre/CBMAT: Copula Based Multivariate Association Test for bivariate mixed phenotypes

#' Simulated mixed bivariate phenotype
#'
#' @description To simulate genotypes, we used the 1000 Genomes Project database.
#' Variants within 500kbs of the BRCA1 gene, for which several known mutations
#' are associated with a higher risk of developing breast, ovarian and prostate
#' cancers, were selected. To avoid any multicollinearity, we pruned variants
#' based on linkage disequilibrium r^2 > 0.7. Further, a total of 503 subjects
#' with a European genetic ancestry are selected in order to avoid any population
#' structure. One discrete and two continuous traits were simulated using a gaussian copula
#' to model the joint dependence. Finally, we also simulated one discrete and one continuous covariate.
#' 
#' @format This data frame has 503 rows and the following 35 columns:
#' \describe{
#'   \item{y.bin}{discrete trait simulated from a latent gaussian variable}
#'   \item{y.gauss}{continous trait simulated from a gaussian distribution}
#'	 \item{y.Gamma}{continous trait simulated from a Gamma distribution}
#'   \item{x1}{intercept}
#'   \item{x2}{discrete covariate}
#'   \item{x3}{continous covariate}
#'   \item{V1:V30}{30 consecutive SNPs sampled from a random genomic region within 500kbs of BRCA1 gene found on chromosome 17}
#' }
#' @source \url{https://www.internationalgenome.org/data/}
"data_mixed"