Learn R Programming

CexoR (version 1.10.0)

CexoR: ChIP-exo peak-pair calling with replicates

Description

ChIP-exo peak-pair calling with replicates.

Usage

cexor(bam, chrN, chrL, p=1e-12, dpeaks=c(0,150), dpairs=100, idr=0.01, N=5e6, bedfile=TRUE, mu=2.6, sigma=1.3, rho=0.8, prop=0.7)

Arguments

bam
BAM alignment files of biological replicates.
chrN
Vector of chromosome names.
chrL
Vector of chromosome sizes (bp).
p
P-value cutoff (should be relaxed to allow the correct estimation of the irreproducible discovery rate (idr). See the vignette for more information.)
dpeaks
Min. and max. allowed distance between peak pairs located at opposed strands in a replicate (bp).
dpairs
Max. allowable distance between peak-pair centres across replicates (bp).
idr
Irreproducible discovery rate cutoff [0-1].
N
Genome is divided in blocks of N bp. for processing. N must be not higher than the size of the smallest chromosome.
bedfile
Generate BED files of ChIP-exo reproducible peak pairs.
mu
A starting value for the mean of the reproducible component (see 'idr' package).
sigma
A starting value for the standard deviation of the reproducible component (see 'idr' package).
rho
A starting value for the correlation coefficient of the reproducible component (see 'idr' package).
prop
A starting value for the proportion of reproducible component (see 'idr' package).

Value

A list containing the following elements:
bindingEvents
A GRanges object with reproducible peak pair locations. The metadata 'value' indicates the Irreproducible discovery rate (IDR) estimated at this region, while 'repI.neg.log10pvalue' indicates -log10(p-value) for the replicate I. 'Stouffer.pvalue' and 'Fisher.pvalue' report the combined p-value considering they come from from independent significance tests.
bindingCentres
A GRanges object with centre position of reproducible peak pair locations. The metadata 'value' indicates the Irreproducible discovery rate (IDR) estimated at this region, while 'repI.neg.log10pvalue' indicates -log10(p-value) for the replicate I. 'Stouffer.pvalue' and 'Fisher.pvalue' report the combined p-value considering they come from from independent significance tests.
pairedPeaksRepl
A GRangesList object with the location of peak pairs retrieved at each replicate. The metadata 'score' indicates -log10(p-value).

Details

Strand specific peak-pair calling in ChIP-exo replicates. The cumulative Skellam distribution function (package 'skellam') is used to detect significant normalized count differences of opposed sign at each DNA strand (peak-pairs). Irreproducible discovery rate for overlapping peak-pairs across biological replicates is estimated using the package 'idr'. The internal functions pskellam and pskellam.sp from the Jerry W. Lewis' 'skellam' R package (version 0.0-8-7) are used to calculate the cumulative Skellam distribution (see LICENSE file).

References

Madrigal P (2015) CexoR: an R/Bioconductor package to uncover high-resolution protein-DNA interactions in ChIP-exo replicates. EMBnet.journal 21: e837.

See Also

CexoR-package

Examples

Run this code

## hg19. chr2:1-1,000,000. CTCF data from Rhee and Pugh (2011)

owd <- setwd(tempdir())

rep1 <- "CTCF_rep1_chr2_1-1e6.bam"
rep2 <- "CTCF_rep2_chr2_1-1e6.bam"
rep3 <- "CTCF_rep3_chr2_1-1e6.bam"
r1 <- system.file("extdata", rep1, package="CexoR",mustWork = TRUE)
r2 <- system.file("extdata", rep2, package="CexoR",mustWork = TRUE)
r3 <- system.file("extdata", rep3, package="CexoR",mustWork = TRUE)

chipexo <- cexor(bam=c(r1,r2,r3), chrN="chr2", chrL=1e6, idr=0.01, p=1e-12, N=3e4)

plotcexor(bam=c(r1,r2,r3), peaks=chipexo, EXT=500)

setwd(owd)

Run the code above in your browser using DataLab