Writes values from input BED files into an in-disk HDF5Array

Source: R/read_beds.R

Writes values from input BED files into an in-disk HDF5Array

read_hdf5_data(
  files,
  ref_cpgs,
  col_list,
  batch_size = 20,
  n_threads = 1,
  h5_temp = NULL,
  zero_based = FALSE,
  strand_collapse = FALSE,
  verbose = TRUE
)

Arguments

files

list of strings; file.paths of BED files
ref_cpgs

data.table; list of CpG sites in the tab-delimited format of chr-start-end. Must be zero-based genome.
col_list

string; The column index object for the input BED files
batch_size

integer; The number of file to hold in memory at once
n_threads

integer; number of threads to use. Default 1. Be-careful - there is a linear increase in memory usage with number of threads. This option is does not work with Windows OS.
h5_temp

string; temporary directory to store hdf5
zero_based

boolean; flag for whether the input data is zero-based or not
strand_collapse

boolean; whether to collapse the crick strand into watson strand. Default FALSE
verbose

boolean; flag to output messages or not.

Value

List of HDF5Array. 1 is methylation, 2 is coverage. If no cov_idx is specified, 2 will be NULL

Details

Using the generated index for genomic coordinates, creates a NA-based dense matrtix of methylation values for each BED file/sample. Each column contains the meth. values for a single sample.

Examples

if (FALSE) {
#Do Nothing
}