mergeLibs: Merge and save libraries of experiment

In Roleren/ORFik: Open Reading Frames in Genomics

Merge and save libraries of experiment

Description

Aggregate count of reads (from the "score" column) by making a merged library. Only allowed for .ofst files!

Usage

mergeLibs(
  df,
  out_dir = file.path(libFolder(df), "ofst_merged"),
  mode = "all",
  type = "ofst",
  keep_all_scores = TRUE,
  paths = filepath(df, type),
  lib_names_full = bamVarName(df, skip.libtype = FALSE),
  max_splits = 20
)

Arguments

df	an ORFik experiment
out_dir	Ouput directory, default file.path(dirname(df$filepath[1]), "ofst_merged"), saved as "all.ofst" in this folder if mode is "all". Use a folder called pshifted_merged, for default Ribo-seq ofst files.
mode	character, default "all". Merge all or "rep" for collapsing replicates only, or "lib" for collapsing all per library type.
type	a character(default: "default"), load files in experiment or some precomputed variant, like "ofst" or "pshifted". These are made with ORFik:::convertLibs(), shiftFootprintsByExperiment(), etc. Can also be custom user made folders inside the experiments bam folder. It acts in a recursive manner with priority: If you state "pshifted", but it does not exist, it checks "ofst". If no .ofst files, it uses "default", which always must exists. Presets are (folder is relative to default lib folder, some types fall back to other formats if folder does not exist): - "default": load the original files for experiment, usually bam. - "ofst": loads ofst files from the ofst folder, relative to lib folder (falls back to default) - "pshifted": loads ofst, wig or bigwig from pshifted folder (falls back to ofst, then default) - "cov": Load covRle objects from cov_RLE folder (fail if not found) - "covl": Load covRleList objects, from cov_RLE_List folder (fail if not found) - "bed": Load bed files, from bed folder (falls back to default) - Other formats must be loaded directly with fimport
keep_all_scores	logical, default TRUE, keep all library scores in the merged file. These score columns are named the libraries full name from bamVarName(df).
paths	character vector, the filpaths to use, default filepath(df, type). Change type argument if not correct. If that is not enough, then you can also update this argument. But be careful about using this directly.
lib_names_full	character vector, default: bamVarName(df, skip.libtype = FALSE). Name to assign to single libraries inside merged file, only kept if mode != "all"
max_splits	integer, default 20. If number of rows to merge > 2^31, how many times can you allow split merging to try to "rescue" the merging process?

Value

NULL, files saved to disc. A data.table with a score column that now contains the sum of scores per merge setting.

Examples

df2 <- ORFik.template.experiment()
df2 <- df2[df2$libtype == "RFP",]
# Merge all
#mergeLibs(df2, tempdir(), mode = "all", type = "default")
# Read as GRanges with mcols
#fimport(file.path(tempdir(), "all.ofst"))
# Read as direct fst data.table
#read_fst(file.path(tempdir(), "all.ofst"))
# Collapse replicates
#mergeLibs(df2, tempdir(), mode = "rep", type = "default")
# Collapse by lib types
#mergeLibs(df2, tempdir(), mode = "lib", type = "default")

Roleren/ORFik documentation built on Nov. 13, 2024, 10 p.m.