Skip to contents

Import splice junction data from BED or SJ.out.tab file

Source: R/jambio-plot.R
import_juncs_from_bed.Rd

Import splice junction data from BED or SJ.out.tab file

Usage

import_juncs_from_bed(
  iBed,
  juncNames,
  sample_id,
  scale_factor = 1,
  use_memoise = FALSE,
  memoise_junction_path = "junctions_memoise",
  gr = NULL,
  verbose = FALSE,
  ...
)

Arguments

iBed

character path or URL to one BED, bigBed, or 'SJ.out.tab' file as produced by STAR during alignment.

  • For 'SJ.out.tab' input, the countUnique value (column 7) is used, with no other filtering applied.

  • For BED and bigBed input, the score is preferred in the name column (column 3), primarily because UCSC tracks limit the score to maximum of 1000.

  • BED name can be numeric, or can be in the form 'JUNC000000_1234' with 'JUNC' prefix followed by integer digits, underscore, the integer values. In this example, the score is '1234'. When not all names can be converted to numeric values, the score column is used.

  • If you see scores with maximum value "1000" the 'score' column could be the problem.

  • bigBed files are imported using the 'cpp11bigwig' package which also uses the appropriate index.

juncNames

character the name of the junction source file

sample_id

character string representing the sample identifier.

scale_factor

numeric default 1, adjusts the raw score, applied by multiplying the scale_factor by each score.

use_memoise

logical default FALSE, whether to cache data using memoise.

memoise_junction_path

character default 'junctions_memoise' with default subdirectory for memoise cache files, used only when use_memoise=TRUE.

gr

GRanges representing the overall range for which junction data will be retrieved. Note that any junctions that span this range, but do not start or end inside this range, will be removed.

verbose

logical indicating whether to print verbose output.

...

additional arguments are ignored.

Details

This function is intended to be called internally by prepareSashimi(), and is provided primarily to enable use of memoise::memoise() to cache results.

This function was refactored in version 0.0.69.900 to handle either BED format, or "SJ.out.tab" junction format as produced by STAR alignment. The method uses data.table::fread() then if there are 9 columns, it assumes the format is "SJ.out.tab". Otherwise it coerces the data.frame with as(bed, "GRanges").

The BED or "SJ.out.tab" file can be gzipped, provided data.table() is able to recognieze and import the compression format.

Note that bigBed format still cannot be used since the rtracklayer package does not support that format.

Also note that junctions with score=0 are dropped at this step, to prevent propagation of junctions with zero counts.

See also

Other Data import functions: getGRcoverageFromBw(), psl2df()