compute_consensus: Compute consensus matrix.

In BEARscc: BEARscc (Bayesian ERCC Assesstment of Robustness of Single Cell Clusters)

Description

Computes the consensus matrix using a data.frame of cluster labels across different BEARscc simulated technical replicates.

Usage

compute_consensus(cluster_labels)

Arguments

cluster_labels

A data.frame of labels assigned to each sample (rownames) across various simulated technical replicates designed by BEARscc (colnames).

Details

We provide a visual and quantitative representation of the clustering variation on a cell-by-cell level by using cluster labels to compute the number of times any given pair of cells associates in the same cluster; this forms the 'noise consensus matrix'. Each element of this matrix represents the fraction of simulated technical replicates in which two cells cluster together (the 'association frequency'), after using a clustering method of the user's choice to generate a data.frame of clustering labels. This consensus matrix may be used to compute BEARscc metrics at both the cluster and cell level.

Value

When the number of samples are n, then the noise consensus resulting from this function is an n x n matrix describing the fraction of simulated technical replicates in which each cell of the experiment associates with another cell.

A brief description of subfunctions

compute_consensus relies on the following subfunction to compute the noise consensus. This function obtains all of the necessary information form the options of compute_consensus.

• names=rownames(cluster_labels)

• create_cm(cluster_labels, names)

Author(s)

David T. Severson <[e-mail]>

Maintainer: Benjamin Schuster-Boeckler <[e-mail]>

See Also

• cluster_consensus()

• report_cluster_metrics()

• report_cell_metrics()

Examples

data("analysis_examples")

noise_consensus <- compute_consensus(clusters.df)
noise_consensus

BEARscc documentation built on Nov. 8, 2020, 7:56 p.m.