qc.samples.cluster: QC sample : clustering

In jmonlong/PopSV: Population-based detection of structural variants from Read-Depth signal

Description

Interactive representation of the bin counts with potential clustering of the samples. This app can be used to check if the reference samples are homogeneous, and define outliers or batches.

Usage

qc.samples.cluster(bc.df, nb.rand.bins = NULL, shiny.app = TRUE)

Arguments

bc.df	a data.frame with the bin counts.
nb.rand.bins	if non-NULL, the number of bins to randomly choose. Default is NULL.
shiny.app	If TRUE (default), a shiny app is launched in a web browser. If FALSE, the function simply returns the PCA results.

Details

There is no need to use all the bin count, a few thousand bins is enough to detect potential problems. 'quick.count' is useful to retrieve bin counts on a subset of bins.

Value

a shiny app in the web browser, or a data.frame with the results of the PCA (if shiny.app=FALSE).

Author(s)

Jean Monlong

jmonlong/PopSV documentation built on Sept. 15, 2019, 9:29 p.m.