Skip to contents

Run CACTI peak-window pipeline genome-wide and add FDR

Source: R/cacti_run_genome.R
cacti_run_genome.Rd

This function is a convenience wrapper that runs the CACTI peak-window pipeline across multiple chromosomes and then computes window-level FDR across all peak windows and chromosomes.

Usage

cacti_run_genome(
  window_size,
  file_pheno_meta,
  file_pheno,
  file_cov,
  chrs,
  qtl_files,
  out_prefix,
  dir_pco = system.file("pco", package = "cacti"),
  min_peaks = 2,
  file_fdr_out = NULL
)

Arguments

window_size

Character like "50kb" (kb units) or numeric (bp).

file_pheno_meta

Path to input phenotype meta table (with header). Columns are -

  • phe_chr: chromosome, with "chr" prefix (e.g. "chr1")

  • phe_from: peak start (integer, < phe_to)

  • phe_to: peak end (integer)

  • phe_id: peak identifier (unique per row)

file_pheno

Path to phenotype expression matrix. Rows = features; first column = feature ID; remaining columns = samples.

file_cov

Path to covariate matrix. Rows = covariates; first column = covariate ID; remaining columns = samples.

chrs

Character vector of chromosome labels (e.g., paste0("chr", 1:22)).

qtl_files

Either:

  • a character vector of the same length as chrs, giving the cis-QTL file path for each chromosome in order; or

  • a single template string containing the placeholder "{chr}", e.g., "extdata/test_qtl_sum_stats_{chr}.txt.gz". In that case, the placeholder is replaced by each element of chrs.

out_prefix

Output prefix used to construct all output filenames.

dir_pco

Directory containing association test helpers: ModifiedPCOMerged_acat.R, liu.R, liumod.R, davies.R, qfc.so.

min_peaks

Minimum number of peaks required in a window to run the multivariate PCO test (>= min_peaks -> PCO; < min_peaks -> univariate p).

file_fdr_out

Optional output path for the FDR-added window-level file. If NULL, a default filename is constructed from out_prefix and window_size.

Value

Invisibly returns a named list of output paths with elements:

file_peak_group

Path to the window-level group.

file_peak_group_peaklevel

Path to the peak-level group.

file_pheno_residual

Path to the residualized phenotype matrix.

file_p_peak_group

Path to the per-window p-value file for all chromosome.

file_fdr_out

Path to the FDR-added window-level result file.

Details

For each chromosome in chrs, it calls cacti_run_chr() and collects the per-window p-value files. It then calls cacti_add_fdr() once, aggregating all chromosomes to obtain q-values for the top-hit p-values in each window.

Examples

if (FALSE) { # \dontrun{
# Example (chr5 as all chr)

file_pheno_meta <- system.file(
  "extdata", "test_pheno_meta.bed",
  package = "cacti"
)

file_pheno <- system.file(
  "extdata", "test_pheno.txt",
  package = "cacti"
)

file_cov <- system.file(
  "extdata", "test_covariates.txt",
  package = "cacti"
)

qtl_file <- system.file(
  "extdata", "test_qtl_sum_stats_chr5.txt.gz",
  package = "cacti"
)

out_prefix <- tempfile("cacti_genome_")

res <- cacti_run_genome(
  window_size = "50kb",
  file_pheno_meta = file_pheno_meta,
  file_pheno = file_pheno,
  file_cov = file_cov,
  chrs = "chr5",
  qtl_files = qtl_file,
  out_prefix = out_prefix,
  dir_pco = system.file("pco", package = "cacti"),
  min_peaks = 2,
  file_fdr_out = file.path(tempdir(), "cacti_fdr_chr5.txt.gz")
)
} # }