manhattan_hicexp: Manhattan plot function for results of multiHiCcompare

In multiHiCcompare: Normalize and detect differences between Hi-C datasets when replicates of each experimental condition are available

Description

Manhattan plot function for results of multiHiCcompare

Usage

manhattan_hicexp(
  hicexp,
  method = "standard",
  return_df = FALSE,
  alpha = 0.05,
  plot.chr = NA
)

Arguments

hicexp	A hicexp object that has had differences detected
method	string denoting the p-value method to use for plotting. Options are "standard", "fisher", "stouffer", "addCLT", and "count". "standard" plots a manhattan plot using all individual p-values (very slow, use with caution). "fisher" or "stouffer" methods use the Fisher's method or the Stouffer-Liptak method, respectively, for combining p-values for each region which are then plotted on the -log10(p-value) Y-axis. "addCLT" combines p-values using the BLMA package's addCLT function. "count" summarizes the number of times a region was detected as significant (see "alpha" parameter), plotted on Y-axis. The higher the dots are, the more sighificant/more frequent a region was detected as significantly differentially interacting.
return_df	Logical, should the data.frame used to generate the plot be returned?
alpha	The adjusted p-value cutoff to be used for calling an interaction significant. This is only used if method = 'count'. Defaults to 0.05.
plot.chr	A numeric value indicating a specific chromosome number to subset the plot to. Defaults to NA indicating that all chromosomes will be plotted.

Details

This function is used to create a manhattan plot for the significance of all genomic regions in the dataset. These correspond to the rows (or columns) of the upper triangle of the full Hi-C matrix. Each genomic region of the Hi-C dataset has multiple interactions it is involved in and the significance of all of these can be visualized with method = "standard". Alternatively the p-values for all these interactions can be combined using either Fisher's method or the Stouffer-Liptak method of combining p-values. Additionally the "count" option will plot based on the number of times each region was found to be involved in a signficantly different interaction. The manhattan plot can be used to identify "hotspot" regions of the genome where major differences seem to be located based on the results of a multiHiCcompare analysis.

Value

A manhattan plot and optionally the data.frame used to generate the manhattan plot.

Examples

data("hicexp_diff")
manhattan_hicexp(hicexp_diff, method = "fisher")

multiHiCcompare documentation built on Nov. 8, 2020, 7:04 p.m.