read_beds: Versatile BedGraph reader.

In CompEpigen/scMethrix: Fast and efficient summarization of single cell whole genome bisulfate data

Description

Versatile BedGraph reader.

Usage

read_beds(
  files,
  ref_cpgs = NULL,
  colData = NULL,
  genome_name = "hg19",
  batch_size = min(20, length(files)),
  n_threads = 1,
  h5 = FALSE,
  h5_dir = NULL,
  h5_temp = NULL,
  desc = NULL,
  verbose = TRUE,
  zero_based = FALSE,
  replace = FALSE,
  fill = TRUE,
  pipeline = c("Custom", "Bismark_cov", "MethylDackel", "MethylcTools", "BisSNP",
    "BSseeker2_CGmap"),
  stranded = FALSE,
  strand_collapse = FALSE,
  chr_idx = NULL,
  start_idx = NULL,
  end_idx = NULL,
  beta_idx = NULL,
  M_idx = NULL,
  U_idx = NULL,
  strand_idx = NULL,
  cov_idx = NULL
)

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.
colData	list of strings; Sample names. Will be derived from filenames if not provided
genome_name	string; Name of genome. Default hg19
batch_size	integer; Max number of files to hold in memory at once. Default 20
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	boolean; Should the coverage and methylation matrices be stored as HDF5Array
h5_dir	string; directory to store H5 based object. This can be NULL and the experiment can be manually saved later
h5_temp	string; temporary directory to store hdf5
desc	string; Description of the experiment
verbose	boolean; flag to output messages or not.
zero_based	boolean; flag for whether the input data is zero-based or not
replace	boolean; flag for whether to delete the contents of h5_dir before saving
fill	boolean; flag whether to fill the output matrixes with all CpGs in ref_cpgs. This must be TRUE for HDF5-based experiments.
pipeline	string; Default NULL. Currently supports "Bismark_cov", "MethylDackel", "MethylcTools", "BisSNP", "BSseeker2_CGmap" If not known use idx arguments for manual column assignments.
stranded	boolean; Whether in input data is stranded. Default FALSE
strand_collapse	boolean; whether to collapse the crick strand into watson strand. Default FALSE
chr_idx	integer; column index for chromosome in bedgraph files
start_idx	integer; column index for start position in bedgraph files
end_idx	integer; column index for end position in bedgraph files
beta_idx	integer; column index for beta values in bedgraph files
M_idx	integer; column index for read counts supporting Methylation in bedgraph files
U_idx	integer; column index for read counts supporting Un-methylation in bedgraph files
strand_idx	integer; column index for strand information in bedgraph files
cov_idx	integer; column index for total-coverage in bedgraph files

Details

Reads BED files and generates methylation matrices. Optionally arrays can be serialized as on-disk HDFS5 arrays.

colData should be input as a headered data.table with a column called "Sample" with names matching the input filenames. Any other columns may be added to include relevant data (e.g. cell type, collection date, etc). During input, this is done as a left join on the inputted files, so the input colData may contain rows for samples that are not actually included in the analysis. This data will be updated on any relevant subsets or merges, etc.

There is an assumption that the first input file will contain the maximum methylation score. It would be extremely unlikely that this assumption is invalid.

Value

An object of class scMethrix

Examples

## Not run: 
#Do Nothing

## End(Not run)

CompEpigen/scMethrix documentation built on Nov. 6, 2021, 3:09 p.m.