correlate_windows: Calculate correlations against top window.

In scFeatureFilter: A correlation-based method for quality filtering of single-cell RNAseq data

Description

Calculates pairwise correlations between all features each window against all features in the reference window.

Usage

correlate_windows(dataset, n_random = 3, ...)

Arguments

dataset	A data frame containing all the binned genes. Usually the output of bin_scdata.
n_random	Number of top window randomization to serve as a negative control. Default to 3.
...	Additional arguments to be passed to cor. Default method is pearson which is the fastest.

Details

This function:

• correlates each feature in each window to each feature in the top window.

• randomize the top window by shuffling expression value, and correlate each gene in each window to the randomized top window. This negative control is repeated as many time as specified by the n_random parameter.

The input of this function is usually the output of the bin_scdata function.

Value

A tibble containing correlation values.

Examples

library(magrittr)
expMat <- matrix(
    c(1, 1, 5,
      1, 2, 3,
      0, 1, 4,
      0, 0, 2),
    ncol = 3, byrow = TRUE, dimnames = list(paste("gene", 1:4), paste("cell", 1:3))
)

calculate_cvs(expMat) %>%
    define_top_genes(window_size = 2) %>%
    bin_scdata(window_number = 1) %>%
    correlate_windows

scFeatureFilter documentation built on Nov. 8, 2020, 7:49 p.m.