Compute the null hypothesis

This function computes the null hypothesis for a given dataset and a list of pathways.

Usage

compute_null_hypothesis(
  root_file_name,
  pathpathways = paste(dirname(root_file_name), "/pathways/", sep = ""),
  pathwaylistfile = paste(dirname(root_file_name), "/pathwaylist.txt", sep = ""),
  nullhypmsfile = paste(dirname(root_file_name), "/nullhyp_ms.txt", sep = ""),
  nullhyps2file = paste(dirname(root_file_name), "/nullhyp_s2.txt", sep = ""),
  seed = 0,
  B = 1,
  alpha = 0.05,
  min_snps = 1,
  max_snps = 15000,
  N_cores = 1,
  n_rows = 1,
  mode = 0,
  implementation = 0,
  verbosity = 0
)

Arguments

root_file_name

full path to a triplet of Plink .bim, .bed and .fam files, without extension

pathpathways

Directory containing one tab-delimited file per SNP-set. Each file must be named <SNP_SET_NAME>.txt, contain no header, and store exactly 3 columns: SNP identifier, chromosome, and base-pair position. Example row: rs123<TAB>17<TAB>43044295. SNP identifiers must match the SNP IDs reported in the input .bim file.

pathwaylistfile

Tab-delimited file describing which SNP-sets to analyse. It must contain one row per SNP-set, with no header, in the form <SNP_SET_NAME><TAB><NUM_SNPS>, where <SNP_SET_NAME> matches the corresponding filename in pathpathways without the .txt extension.

nullhypmsfile

name of the file where to write the estimated confidence thresholds for each pathway

nullhyps2file

name of the file where to write the estimated confidence thresholds for each pathway

seed

seed of the random number generator (default 0)

B

Number of bootstrap replicates. Use -1 to enable automatic bootstrap selection.

alpha

threshold

min_snps

You must provide a number of minimum SNPs greater than 0.

max_snps

You must provide a number of maximum SNPs greater than 0.

N_cores

Number of cores to use

n_rows

Number of rows to use

mode

Compatibility parameter. Only 0 (discrete phenotype) is supported.

verbosity

Integer verbosity level (0 = minimal output, higher values increase logging).

Details

This function computes the null hypothesis for a given dataset and a list of pathways. The null hypothesis is computed by bootstrapping the dataset and computing the p-values for each pathway. The p-values are then used to estimate the confidence thresholds for each pathway.

Examples

if (FALSE) { # \dontrun{
root <- tempfile("synthetic")
synthetic <- nebula::simulate_synthetic_genotype(
    n_snps = 20,
    n_individuals = 20,
    n_causal_snps = 3
)
nebula::save_synthetic_genotype(synthetic, file = root)
nebula::generate_synthetic_snpsets(
    input_file = paste0(root, ".bim"),
    snps_per_set = 5,
    num_sets = 4,
    root_dir = dirname(root)
)
} # }