Package 'PeakSegDP'

calc grad list

Description

List of calc.grad functions: x, features, limits -> gradient.

Usage

"calc.grad.list"

---

calc loss from lp list

Description

if we have already calculated the linear predictor using fit$predict, this function can be useful.

Usage

"calc.loss.from.lp.list"

---

calc loss list

Description

List of interval regression loss functions: x, feat, lim => numeric.

Usage

"calc.loss.list"

---

cDPA

Description

A constrained dynamic programming algorithm (cDPA) can be used to compute the best segmentation with respect to the Poisson likelihood, subject to a constraint on the number of segments, and the changes which must alternate: up, down, up, down, ...

Usage

cDPA(count, weight = rep(1, 
    length(count)), maxSegments)

Arguments

count	Integer vector of count data to segment.
weight	Data weights (normally this is the number of base pairs).
maxSegments	Maximum number of segments to consider.

Author(s)

Toby Dylan Hocking, Guillem Rigaill

Examples

fit <- cDPA(c(0, 10, 11, 1), maxSegments=3)
stopifnot(fit$ends[3,4] == 3)
stopifnot(fit$ends[2,3] == 1)

---

ChIP-seq aligned read coverage for 4 samples on a subset of chr11

Description

A ChIP-seq experiment was performed to locate the genomic positions of a histone (H3K4me3) in 2 B cell samples (McGill0091, McGill0322) and 2 T cell samples (McGill0002, McGill0004). The short sequence reads (about 100 base pairs each) were aligned to the hg19 reference genome, and the "coverage" in this data set contains the total count of aligned reads at each base pair. It also contains annotated regions determined by an expert who examined scatterplots of the coverage profiles.

Usage

data("chr11ChIPseq")

Format

A named list of 2 data.frames: regions contains annotations about which regions contain or do not contain peaks, and coverage contains the noisy signal.

Source

H3K4me3_TDH_immune chunk 5 in http://cbio.ensmp.fr/~thocking/chip-seq-chunk-db/ which in turn comes from http://epigenomesportal.ca/

Examples

data(chr11ChIPseq)
library(ggplot2)
ann.colors <-
  c(noPeaks="#f6f4bf",
    peakStart="#ffafaf",
    peakEnd="#ff4c4c",
    peaks="#a445ee")

if(interactive() && require(ggplot2)){

ggplot()+
  scale_fill_manual("annotation", values=ann.colors,
                    breaks=names(ann.colors))+
  penaltyLearning::geom_tallrect(aes(xmin=chromStart/1e3, xmax=chromEnd/1e3,
                    fill=annotation),
                data=chr11ChIPseq$regions, alpha=1/2)+
  theme_bw()+
  theme(panel.margin=grid::unit(0, "cm"))+
  facet_grid(sample.id ~ ., scales="free")+
  geom_step(aes(chromStart/1e3, count), data=chr11ChIPseq$coverage)+
  xlab("position on chr11 (kilo base pairs)")

}

---

Counts of first base of aligned reads

Description

For 4 samples on chr11 (hg19), this data set counts the first base pair of aligned reads at each genomic position. In contrast, chr11ChIPseq counts every base pair in each read (and each read is about 100bp, so that means there is some auto-correlation in chr11ChIPseq, but not in chr11first).

Usage

data("chr11first")

Format

A data frame with 23252 observations on the following 4 variables.

sample.id

a factor with levels for each of 4 samples

chromStart

integer vector: base before, on chr11

chromEnd

integer vector: last base on chr11

count

integer: aligned first base read counts

Source

H3K4me3_TDH_immune chunk 5 in http://cbio.ensmp.fr/~thocking/chip-seq-chunk-db/ which in turn comes from http://epigenomesportal.ca/

Examples

data(chr11ChIPseq)
data(chr11first)
library(ggplot2)
ann.colors <-
  c(noPeaks="#f6f4bf",
    peakStart="#ffafaf",
    peakEnd="#ff4c4c",
    peaks="#a445ee")
both <- list(coverage=chr11ChIPseq$coverage, first=chr11first)
representations <- NULL
one.sample <- "McGill0322"
for(data.type in names(both)){
  one <- subset(both[[data.type]], sample.id==one.sample)
  representations <- rbind(representations, data.frame(data.type, one))
}
one.sample.regions <- subset(
  chr11ChIPseq$regions, sample.id==one.sample)

if(interactive() && require(ggplot2)){

ggplot()+
  scale_fill_manual("annotation", values=ann.colors,
                    breaks=names(ann.colors))+
  penaltyLearning::geom_tallrect(aes(xmin=chromStart/1e3, xmax=chromEnd/1e3,
                    fill=annotation),
                data=one.sample.regions, alpha=1/2)+
  theme_bw()+
  theme(panel.margin=grid::unit(0, "cm"))+
  facet_grid(data.type ~ ., scales="free")+
  geom_step(aes(chromStart/1e3, count), data=representations)+
  xlab("position on chr11 (kilo base pairs)")

}

---

derivs

Description

List of functions, each a derivative of a phi loss.

Usage

"derivs"

---

GeomTallRect

Description

ggproto object for geom_tallrect

Usage

"GeomTallRect"

---

getPath

Description

Extract endpoint matrix from cDPA result.

Usage

getPath(A)

Arguments

A

A

Author(s)

Toby Dylan Hocking, Guillem Rigaill

---

Several ChIP-seq profiles, some of which have few data points

Description

These data are used to test the PeakSegDP algorithm, to make sure it gives sensible results, even when there are few data.

Usage

data("H3K36me3.AM.immune.19")

Format

Named list of 21 data.frames, each with columns chromStart, chromEnd, count.

Source

http://cbio.ensmp.fr/~thocking/chip-seq-chunk-db/ data set H3K36me3_AM_immune, chunk id 19

---

8 profiles of H3K36me3 data

Description

these data caused a bug in multiSampleSegHeuristic.

Usage

data("H3K36me3.TDH.other.chunk3.cluster4")

Format

A data frame with 36914 observations on the following 4 variables.

sample.id

a factor with 8 levels

chromStart

integer vector

chromEnd

integer vector

count

integer vector

Source

http://cbio.ensmp.fr/~thocking/chip-seq-chunk-db/ data set H3K36me3_TDH_other chunk 3.

---

Histone ChIP-seq data, 26 samples, chr1 subset

Description

26 samples, each with the same overlapping peak(s).

Usage

data("H3K4me3.TDH.immune.chunk12.cluster4")

Format

A data frame.

Source

http://cbio.ensmp.fr/~thocking/chip-seq-chunk-db H3K4me3_TDH_immune data set, chunk.id=12.

---

PeakSegDP

Description

Compute the PeakSeg model on a data.frame of compressed sequence reads.

Usage

PeakSegDP(compressed, 
    maxPeaks)

Arguments

compressed	data.frame with columns chromStart, chromEnd, count.
maxPeaks	maximum number of peaks to consider.

Author(s)

Toby Dylan Hocking, Guillem Rigaill

Examples

library(PeakSegDP)
data(chr11ChIPseq, envir=environment())
one <- subset(chr11ChIPseq$coverage, sample.id=="McGill0002")[10000:12000,]
fit <- PeakSegDP(one, 3L)

if(interactive() && require(ggplot2)){

  ggplot()+
    geom_step(aes(chromStart/1e3, count), data=one)+
    geom_segment(aes(chromStart/1e3, mean,
                     xend=chromEnd/1e3, yend=mean),
                 data=fit$segments, color="green")+
    geom_segment(aes(chromStart/1e3, 0,
                     xend=chromEnd/1e3, yend=0),
                 data=subset(fit$segments, status=="peak"),
                 size=3, color="deepskyblue")+
    theme_bw()+
    theme(panel.margin=grid::unit(0, "cm"))+
    facet_grid(peaks ~ ., scales="free", labeller=function(df){
      s <- ifelse(df$peaks==1, "", "s")
      df$peaks <- paste0(df$peaks, " peak", s)
      df
    })

}

---

phi list

Description

List of functions, each a phi loss.

Usage

"phi.list"

---

PoissonLoss

Description

Compute the weighted Poisson loss function, which is seg.mean - count * log(seg.mean). The edge case is when the mean is zero, in which case the probability mass function takes a value of 1 when the data is 0 (and 0 otherwise). Thus the log-likelihood of a maximum likelihood segment with mean zero must be zero.

Usage

PoissonLoss(count, seg.mean, 
    weight = 1)

Arguments

count	count
seg.mean	seg.mean
weight	weight

Author(s)

Toby Dylan Hocking, Guillem Rigaill

Examples

PoissonLoss(1, 1)
PoissonLoss(0, 0)
PoissonLoss(1, 0)
PoissonLoss(0, 1)

---

regression funs

Description

List of regression functions: features, limits -> list.

Usage

"regression.funs"

---