R/build_cross2_dataset.R

In pujana-lab/systematicQTL: Pipeline to run sistematically QTL analysis

#' build_cross2_dataset
#'
#' Generates cross2 dataset from basic geno and phenotypes name. Geno file should combine genenotype values and mapping (see to_csvs_geno) and phenotypes table should be a matrix-type object (users in rows, phenotypes in columns).
#'  This function is the entry point to QtlWrapper object.
#'
#' @param geno_filename Path to csvs geno file
#' @param pheno_filename Path to csv pheno file
#' @param covar_column_names List of "phenotype" columns related to adjusting covariates.
#' @param pheno_column_names List of "phenotype" columns realted to main variables to study.
#' @param genotypes List of allowed genotype values of genotypes file
#' @param alleles  List of homozygote alleles included, should be a subset of genotypes variable
#' @param na.strings List of strings associated to NA values
#' @param comment.char List or string of allowed comment chars
#' @param ... Additional arguments passed to read.cross function
#'
#' @return This function returns an R/QTL Object of class "cross2". For details, see the \href{https://kbroman.org/qtl2/assets/vignettes/developer_guide.html}{R/QTL2} developer guide.
#'
#'
#' @examples
#' \dontrun{
#'   geno_data_matrix_filename = '/path/to/geno_file'
#'   geno_map_filename = '/path/to/geno_mapping_file'
#'   phenotype_filenam = '/path/to/phenotypes_file'
#'
#'   genotypes_csvs  = to_csvs_geno(geno_map_filename, geno_data_matrix_filename)
#'   write_csvs_geno(genotypes_csvs, '/path_to_geno_csvs')
#'
#'   chosen_genotypes    = c('A', 'H', 'B')
#'   chosen_alleles      = c('A','B')
#'   chosen_covars       = c('pheno_var1', 'pheno_var2', 'pheno_var3')
#'   chosen_phenotypes   = c('signature1', 'signature2', 'signature3')
#'
#'   cross2 = systematicQTL::build_cross2_dataset(
#'     geno_filename = geno_filename,
#'     pheno_filename = pheno_filename,
#'     covar_column_names = chosen_covars,
#'     pheno_column_names = chosen_phenotypes,
#'     alleles = chosen_alleles
#'   )
#'
#' }
#'
#'
#' @export
#'
build_cross2_dataset <- function(
  geno_filename,
  pheno_filename,
  covar_column_names,
  pheno_column_names,
  genotypes = c('A', 'H','B'),
  alleles  = c('A', 'B'),
  na.strings = c('-1', 'discrepancy', 'NA'),
  comment.char = '#',
  ...
) {

   dataset = qtl::read.cross(
    'csvs',
    genfile = geno_filename,
    phefile = pheno_filename,
    genotypes = genotypes,
    alleles = alleles,
    na.strings = na.strings,
    comment.char = comment.char,
    convertXdata = FALSE,
    ...
  )

  dataset = qtl::jittermap(dataset)

  dataset = qtl2::convert2cross2(dataset)

  if(is.null(dataset$covar)){
    dataset$covar = data.frame(row.names = rownames(dataset$pheno))
  }

  for(covariate in covar_column_names){
    dataset$covar[,covariate] = dataset$pheno[,covariate]
  }

  dataset$pheno = dataset$pheno[,colnames(dataset$pheno) %in% pheno_column_names]

  return(dataset)

}