plotDI: Construct a plaid plot of differential interactions
In diffHic: Differential Analyis of Hi-C Data

Description Usage Arguments Details Value Author(s) References See Also Examples

Plot differential interactions in a plaid format with informative coloring.

plotDI(data, fc, first.region, second.region=first.region, col.up="red", 
    col.down="blue", background="grey70", zlim=NULL, xlab=NULL, ylab=NULL, 
    diag=TRUE, ...)

rotDI(data, fc, region, col.up="red", col.down="blue", background="grey70", 
    zlim=NULL, xlab=NULL, max.height=NULL, ylab="Gap", ...)

`data`	an InteractionSet object
`fc`	a numeric vector of log-fold changes
`first.region`	a `GRanges` object of length 1 specifying the first region
`second.region`	a `GRanges` object of length 1 specifying the second region
`region`	a `GRanges` object of length 1 specifying the region of interest
`col.up`	any type of R color to describe the maximum color for positive log-fold changes
`col.down`	any type of R color to describe the maximum color for negative log-fold changes
`background`	any type of R color, specifying the background color of the interaction space
`zlim`	a numeric scalar indicating the maximum absolute log-fold change
`xlab`	character string for the x-axis label on the plot, defaults to the first chromosome name
`max.height`	a numeric scalar indicating the y-axis limit for `rotPlaid`
`ylab`	character string for the y-axis label on the plot, defaults to the second chromosome name in `plotDI`
`diag`	a logical scalar specifying whether boxes should be shown above the diagonal for intra-chromosomal plots in `plotDI`
`...`	other named arguments to be passed to `plot`

The plotDI function constructs a plaid plot on the current graphics device. The intervals of first.region and second.region are represented by the x- and y-axes, respectively. Each bin pair is represented by a box in the plotting space, where each side of the box represents a bin. Plotting space that is not covered by any bin pair is shown in background.

The color of the box depends on the magnitude and sign of the log-fold change in fc. Positive log-FCs will range from white to col.up, whereas negative log-FCs will range from white to col.down. The chosen color is proportional to the magnitude of the log-FC, and the most extreme colors are only obtained at the maximum absolute log-FC in fc. The maximum value can be capped at zlim for better resolution of small log-FCs.

If diag=TRUE, boxes will also be plotted above the diagonal for intra-chromosomal plots. This is set as the default to avoid confusion when first.region is not set as the anchor range, i.e., it has a lower sorting order than second.region. However, this can also be turned off to reduce redundancy in visualization around the diagonal.

The rotDI function constructs a rotated plot of differential interactions, for visualization of local changes. See rotPlaid for more details.

A (rotated) plaid plot of differential interactions is produced on the current graphics device. A function is also invisibly returned that converts log-FCs into colors. This is useful for coordinating the colors, e.g., when constructing a separate color bar.

Aaron Lun

Lieberman-Aiden E et al. (2009). Comprehensive Mapping of Long-Range Interactions Reveals Folding Principles of the Human Genome. Science 326, 289-293.

plotPlaid, rotPlaid, squareCounts

# Setting up the objects.
a <- 10
b <- 20
regions <- GRanges(rep(c("chrA", "chrB"), c(a, b)), IRanges(c(1:a, 1:b), c(1:a, 1:b)),
    seqinfo=Seqinfo(seqlengths=c(chrA=a, chrB=b), seqnames=c("chrA", "chrB")))

set.seed(3423)
all.anchor1 <- sample(length(regions), 500, replace=TRUE)
all.anchor2 <- as.integer(runif(500, 1, all.anchor1+1))
out <- InteractionSet(matrix(0, 500, 1), colData=DataFrame(lib.size=1000),
        GInteractions(anchor1=all.anchor1, anchor2=all.anchor2, 
            regions=regions, mode="reverse"), metadata=List(width=1))
fc <- runif(nrow(out), -2, 2)

# Constructing intra-chromosomal DI plots around various regions
plotDI(out, fc, first.region=GRanges("chrA", IRanges(1, 10)), 
    second.region=GRanges("chrA", IRanges(1, 10)), diag=TRUE)
plotDI(out, fc, first.region=GRanges("chrA", IRanges(1, 10)), 
    second.region=GRanges("chrA", IRanges(1, 10)), diag=FALSE)

# Constructing inter-chromosomal DI plots around various regions
xxx <- plotDI(out, fc, first.region=GRanges("chrB", IRanges(1, 10)), 
    second.region=GRanges("chrA", IRanges(1, 20)), diag=TRUE)
plotDI(out, fc, first.region=GRanges("chrB", IRanges(1, 100)), 
    second.region=GRanges("chrA", IRanges(1, 200)), diag=TRUE, zlim=5)

# Making colorbars.
xxx((-10):10/10)
xxx((-20):20/20)

# Rotated.
rotDI(out, fc, region=GRanges("chrA", IRanges(1, 200)))
rotDI(out, fc, region=GRanges("chrB", IRanges(1, 200)))