plotCor: Plot the correlation between ChIP-seq experiments

Description Usage Arguments Details Value Methods (by class) Author(s) See Also Examples

Description

Plot the correlation between ChIP-seq experiments using heatscatter plot or, if heatscatterplot = FALSE, correlation tables.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
  plotCor(theObject, rawFile = FALSE, rpmFile = FALSE, bgsubFile = FALSE, 
revFile = FALSE, spiked = TRUE, main = "", add_contour = FALSE, method_cor = 
"spearman", nlevels = 10, color_contour = "black", show_cor = TRUE, 
allOnPanel = TRUE, method_scale = "none", method_corrplot = "circle", 
heatscatterplot = TRUE, type_corrplot = "upper", diag_corrplot = FALSE, 
separateWindows = FALSE, verbose = FALSE, ...)

## S4 method for signature 'ChIPSeqSpikeDataset'
plotCor(theObject, rawFile = FALSE, rpmFile = 
FALSE, bgsubFile = FALSE, revFile = FALSE, spiked = TRUE, main = "", 
add_contour = FALSE, method_cor = "spearman", nlevels = 10, color_contour = 
"black", show_cor = TRUE, allOnPanel = TRUE, method_scale = "none", 
method_corrplot = "circle", heatscatterplot = TRUE, type_corrplot = "upper", 
diag_corrplot = FALSE, separateWindows = FALSE, verbose = FALSE, ...)

## S4 method for signature 'ChIPSeqSpikeDatasetBoost'
plotCor(theObject, rawFile = FALSE, 
rpmFile = FALSE, bgsubFile = FALSE, revFile = FALSE, spiked = TRUE, main = "", 
add_contour = FALSE, method_cor = "spearman", nlevels = 10, 
color_contour = "black", show_cor = TRUE, allOnPanel = TRUE, method_scale = 
"none", method_corrplot = "circle", heatscatterplot = TRUE, type_corrplot = 
"upper", diag_corrplot = FALSE, separateWindows = FALSE, verbose = FALSE, ...)

## S4 method for signature 'ChIPSeqSpikeDatasetList'
plotCor(theObject, rawFile = FALSE, rpmFile
= FALSE, bgsubFile = FALSE, revFile = FALSE, spiked = TRUE, main = "", 
add_contour = FALSE, method_cor = "spearman", nlevels = 10, color_contour = 
"black", show_cor = TRUE, allOnPanel = TRUE, method_scale = "none",
 method_corrplot = "circle", heatscatterplot = TRUE, type_corrplot = "upper", 
 diag_corrplot = FALSE, separateWindows = FALSE, verbose = FALSE, ...)

## S4 method for signature 'ChIPSeqSpikeDatasetListBoost'
plotCor(theObject, rawFile = FALSE, 
rpmFile = FALSE, bgsubFile = FALSE, revFile = FALSE, spiked = TRUE, main = "", 
add_contour = FALSE, method_cor = "spearman", nlevels = 10, 
color_contour = "black", show_cor = TRUE, allOnPanel = TRUE, 
method_scale = "none", method_corrplot = "circle", heatscatterplot = TRUE, 
type_corrplot = "upper", diag_corrplot = FALSE, separateWindows = FALSE, 
verbose = FALSE, ...)

Arguments

theObject

ChIPSeqSpike dataset (see ?spikeDataset)

rawFile

If TRUE, use the untransformed data. Not available in boost mode. Default is FALSE. (see details)

rpmFile

If TRUE, use the RPM scaled data. Not available in boost mode. Default is FALSE. (see details)

bgsubFile

If TRUE, use the input subtracted data. Not available in boost mode. Default is FALSE. (see details)

revFile

If TRUE, use the RPM reverted data. Not available in boost mode. Default is FALSE. (see details)

spiked

If TRUE, use the spiked data. Default is TRUE. (see details)

main

Main title of the plot. No title is displayed by default.

add_contour

If TRUE, contours are added to the heatscatter. Default is FALSE.

method_cor

A character string indicating which correlation coefficient is to be computed. One of 'pearson' (default) , 'kendall' or 'spearman'.

nlevels

An integer giving the number of levels of the contour lines. Not used if heatscatteplot is FALSE. Default is 10

color_contour

Character string defining the color of the contour line. Not used if heatscatteplot is FALSE. Default is 'black'.

show_cor

Logical indicating if the correlation is added to the title. Not used if heatscatteplot is FALSE. Default is TRUE.

allOnPanel

Logical indicating if all correlations should be on the same panel. Not used if heatscatteplot is FALSE. Default is TRUE.

method_scale

Character string indicating the scaling to be applied to the data. Possible values are 'none', 'log', 'asinh', 'cuberoot' or 'zscore'. Not used if heatscatteplot is FALSE.

method_corrplot

If heatscatterplot is FALSE, define the graphical representation used for the correlation table. Possible values are 'circle', 'square', 'ellipse', 'number', 'pie', 'shade' and 'color'. See ?corrplot::corrplot for more details. Default is 'circle'.

heatscatterplot

If TRUE, use a heatscatter representation instead of correlation table. see ?LSD::heatscatter for more details. Default is TRUE.

type_corrplot

If heatscatter is FALSE, define if the full ('full'), the lower triangular ('lower') or upper triangular matrix is displayed. Default is 'upper'.

diag_corrplot

If heatscatter is FALSE, logical indicating if the correlation coefficients are displayed on the principal diagonal. Default is FALSE.

separateWindows

If heatscatterplot is TRUE, Logical indicating if each plot is output to a separate window. Default is FALSE.

verbose

Logical indicating if processing messages are displayed. Default is FALSE

...

Additional parameter to pass to the LSD::heatscatter or corrplot::corrplot functions.

Details

The 'rawFile', 'rpmFile', 'bgsubFile', 'revFile', 'spiked' parameters indicate if the untransformed, RPM scaled, input DNA subtracted, RPM reversed or spiked data should be used plotted. This option is only available if not in boost mode (see ?spikePipe for details on what these steps are).

Value

If heatscatterplot is FALSE, return the correlation matrix.

Methods (by class)

Author(s)

Nicolas Descostes

See Also

spikeDataset heatscatter corrplot spikePipe plotTransform boxplotSpike plotHeatmaps plotProfile

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
## Sub-sample
data("result_extractBinding")

## Heatscatter of spiked data using Spearman correlation
plotCor(csds)

## Pearson correlation of log transformed raw data
plotCor(csds,  rawFile = TRUE, spiked = FALSE,  main = "heatscatter",  
method_cor = "pearson", method_scale = "log")
 
## Correlation table of all transformation steps with circle representation
plotCor(csds, rawFile = TRUE, rpmFile = TRUE, bgsubFile = TRUE, revFile = TRUE,
 spiked = TRUE, heatscatterplot = FALSE, verbose = TRUE)
 
## Correlation table of all transformation steps with number representation
plotCor(csds, rawFile = TRUE, rpmFile = TRUE, bgsubFile = TRUE, revFile = TRUE,
 spiked = TRUE, heatscatterplot = FALSE, verbose = TRUE, method_corrplot = 
 "number")

descostesn/ChIPSeqSpike documentation built on Aug. 6, 2019, 3:51 p.m.