getCorrelation-methods: get correlation between samples in methylBase or methylBaseDB...

In methylKit: DNA methylation analysis from high-throughput bisulfite sequencing results

Description

The functions returns a matrix of correlation coefficients and/or a set of scatterplots showing the relationship between samples. The scatterplots will contain also fitted lines using lm() for linear regression and lowess for polynomial regression.

Usage

getCorrelation(object,method="pearson",plot=FALSE,nrow)

## S4 method for signature 'methylBase'
getCorrelation(object, method = c("pearson", "kendall", "spearman"), plot)

## S4 method for signature 'methylBaseDB'
getCorrelation(object, method = "pearson", plot = FALSE, nrow = 2e+06)

Arguments

object	a methylBase or methylBaseDB object
method	a character string indicating which correlation coefficient (or covariance) is to be computed (default:"pearson", other options are "kendall" and "spearman")
plot	scatterPlot if TRUE (default:FALSE)
nrow	a numeric giving the number of lines to read in of methylBaseDB object, defaults to 2e6

Value

a correlation matrix object and plot scatterPlot

Details

The argument 'nrow' is only evaluated if the input is a methylBaseDB object. If 'nrow' is not specified getCorrelation will read the first 2M records of the given object, but if you want to read all records 'nrow' has to be NULL. You should change 'nrow' if using getCorrelation with all records of the methylBaseDB object would take too long.

If the scatter plot is plotted, the red line in the plot is from linear regression fit and the green line is from polynomial regression fit with stats::lowess.

Examples

data(methylKit)

getCorrelation(methylBase.obj,method="pearson",plot=FALSE)

# create methylBaseDB
methylBaseDB.obj <- unite(methylRawList.obj,save.db=TRUE,dbdir="methylDB")

getCorrelation(methylBaseDB.obj,method="pearson",plot=FALSE,nrow=10000)




# remove Database again
rm(methylBaseDB.obj)
unlink("methylDB",recursive=TRUE)

methylKit documentation built on Jan. 30, 2021, 2 a.m.