detectTSS: Detection of Trancription start sites based on local...
In vivekbhr/icetea: Integrating Cap Enrichment with Transcript Expression Analysis

Description Usage Arguments Value Examples

Detection of Trancription start sites based on local enrichment

detectTSS(
  CSobject,
  groups,
  outfile_prefix = NULL,
  windowSize = 10L,
  sliding = TRUE,
  foldChange = 2,
  mergeLength = 1L,
  restrictChr = NULL,
  ncores = 1,
  readPos = "start"
)

## S4 method for signature 'CapSet'
detectTSS(
  CSobject,
  groups,
  outfile_prefix = NULL,
  windowSize = 10L,
  sliding = TRUE,
  foldChange = 2,
  mergeLength = 1L,
  restrictChr = NULL,
  ncores = 1,
  readPos = "start"
)

`CSobject`	CapSet object created using `newCapSet` function
`groups`	a character vector that contains group name of the sample, for replicate-based TSS calling (see example)
`outfile_prefix`	Output name prefix for the .Rdata file containing window counts, background counts and filtering statistics calculated during TSS detection.
`windowSize`	Size of the window to bin the genome for TSS detection. By default, a window size of 10 is used for binning the genome, however smaller window sizes can optionally be provided for higher resolution TSS detection. Note that the background size is set to 200x the window size (2kb for 10bp windows) to calculate local enrichment. Subsequently enriched windows are merged, unless the mergeLength is increased.
`sliding`	TRUE/FALSE. Indicating whether or not to use sliding windows. The windows are shifted by length which is half of the specified window length.
`foldChange`	Numeric. A fold change cutoff of local enrichment to detect the TSS. If the samples have good signal enrichment over background (inspect in genome browser), a low cutoff of 2-fold can be used. For samples with low sequencing depth it's also desirable to have a low cutoff of 2-fold. The final "score" of detected TSS is the mean fold-change of all merged windows that passed the foldChange cutoff. TSSs can therefore also be filtered using this score after detectTSS is run.
`mergeLength`	Integer. Merge the windows within this distance that pass the foldChange cutoff. Default (1L) means that only subsequently enriched windows would be merged.
`restrictChr`	Chromosomes to restrict the analysis to.
`ncores`	No. of cores/threads to use
`readPos`	character. position of read to use. Options are "start", "end" and "center". For TSS detection, the "start" of reads are used (default). But center or end might be useful for detecting RNA-binding proteins (in iCLIP-like data)

.bed files containing TSS position for each group, along with a bed file for consensus (union) TSS sites of all samples.

# before running this
# 1. Create a CapSet object
# 2. de-multiplex the fastqs
# 3. map them
# 4. filter duplicate reads from mapped BAM

# load a previously saved CapSet object
cs <- exampleCSobject()
# detect TSS (samples in same group are treated as replicates)
cs <- detectTSS(cs, groups = rep(c("wt","mut"), each = 2), outfile_prefix = "testTSS",
           foldChange = 6, restrictChr = "X", ncores = 1)