normalize_chrom: Normalize an InteractionSet with iterative proportional...

In AnnikaGable/chromium: Analyze and visualize high throughput chromatin interaction data, such as Hi-C or ChIA-PET

Description

A binned or unbinned InteractionSet is normalized using iterative proportional fitting (IPF), an implementation of the iterative matrix balancing algorithm by Pukelsheim and Simeone (2009).

Usage

normalize_chrom(Iset, numberOfIterations = 40L, plotLoss = FALSE)

Arguments

Iset	An InteractionSet with binned or unbinned data. If the data is binned, metadata(Iset)$binSize must contain the bin size.
numberOfIterations	The number of iterations for normalization. Usually, 20 should be more than enough.
plotLoss	Optional boolean parameter, to determine if the loss function should be plotted. Defaults to FALSE.

Value

A list containing the normalized InteractionSet, i.e. assay(normIset) will contain the normalized signal. In the metadata, the number of interchromosomal contacts and intrachromosomal contacts is added, as well as the contact distances, i.e. the distances between the genomic positions where two interacting bins or restriction fragments start. In addition, the loss function is given for every iteration of the normalization. If plotLoss = T, the metadata will also contain a plotted version of the loss function, which can be displayed with metadata(normIset)$lossPlot.

Examples

#norm_Iset <- normalize_chrom(raw_Iset, numberOfIterations = 40L, plotLoss = FALSE)

AnnikaGable/chromium documentation built on May 5, 2019, 6:04 a.m.