Calculate PRS using plink2

Calculate PRS using plink2

Usage

plink_prs(
  base_df,
  id_col,
  effect_allele_col,
  effect_size_col,
  multithread = TRUE,
  pfile_dir,
  plink_bin
)

Arguments

base_df

(dataframe) A dataframe containing at minimum variant ids, effect alleles, and corresponding effect sizes

id_col

(character) Name of column containing variant ids, corresponding to the variant ids in the plink files

effect_allele_col

(character) Name of column containing effect alleles

effect_size_col

(character) Name of column containing effect estimates

multithread

(logical) Should multi-threading be used to compute PRS across all pfiles in parallel (default = TRUE)

pfile_dir

(character) Directory containing the plink files for the target dataset. If the directory contains multiple files (for example, one per chromosome), the PRS can be calculated in parallel.

plink_bin

(character) Path to executable plink2 binary

Value

A tibble containing rows for each individual in the target dataset, and a column containing their PRS

Examples

if (FALSE) {
# load weights file into dataframe
weights_df <- fs::path_package("extdata", "PGS000667-PMBB.txt", package = "prsr") %>%
  vroom::vroom()

# calculate prs
prs_res <- plink_prs(
  base_df = weights_df,
  id_col = variant,
  effect_allele_col = effect_allele,
  effect_size_col = effect_weight_mg_dl,
  pfile_dir = "/project/PMBB/PMBB-Release-2020-2.0/Imputed/pgen/",
  plink_bin = "/scratch/Applications/PLINK/plink2_linux_x86_64_20211125/plink2a"
)
}