swinsor: Smooth Winsorization

In RNAprobR: An R package for analysis of massive parallel sequencing based RNA structure probing data

Description

Performs sliding window Winsorization given treated GRanges generated by comp() function. It winsorizes values in windows (of a size specified by window_size) sliding by 1 nt over whole transcript length and reports mean winsorized value for each nucleotide (as well as standard deviation).

Usage

swinsor(Comp_GR, winsor_level = 0.9, window_size = 71, only_top = FALSE,
  nt_offset = 1, add_to)

Arguments

Comp_GR	GRanges object made by comp() function.
winsor_level	Winsorization level. Bottom outliers will be set to (1-winsor_level)/2 quantile and top outliers to (1+winsor_level)/2 quantile.
window_size	Size of a sliding window.
only_top	If TRUE then bottom values are not Winsorized and are set to 0.
nt_offset	How many position in the 5' direction should the signal be offset to account for the fact that reverse transcription termination occurs before site of modification.
add_to	GRanges object made by other normalization function (dtcr(), slograt(), swinsor(), compdata()) to which normalized values should be added.

Value

GRanges object with "swinsor" (mean smooth-Winsor values) and "swinsor.sd" (standard deviation of smooth-Winsor values) metadata.

Author(s)

Lukasz Jan Kielpinski, Jeppe Vinther, Nikos Sidiropoulos

References

"Analysis of sequencing based RNA structure probing data" Kielpinski, Sidiropoulos, Vinther. Chapter in "Methods in Enzymology" (in preparation)

See Also

comp, dtcr, slograt, compdata, GR2norm_df, plotRNA, norm2bedgraph, winsor, swinsor_vector

Examples

dummy_euc_GR <- GRanges(seqnames="DummyRNA",
                        IRanges(start=round(runif(100)*100),
                        width=round(runif(100)*100+1)), strand="+",
                        EUC=round(runif(100)*100))
dummy_comp_GR <- comp(dummy_euc_GR)
swinsor(dummy_comp_GR)

RNAprobR documentation built on Nov. 8, 2020, 5:57 p.m.