get_marker_genes: Get marker genes for each cluster

In shibiaowan/SHARP: hyperfast and accurate single-cell RNA-Seq data clustering

Description

This is to identify marker genes for each cluster after SHARP clustering. It uses a method similar to that proposed in the SC3 package except two improvements: (1) it uses an adjusted threshold to select marker genes instead of using a hardthreshod (i.e., p-value < 0.01 and AUROC > 0.85), so that marker genes can be found for all clusters; (2) it uses a parallelization way to calculate the p-value and AUROC (areas under the curve of the Receiver Operating Characteristic) for each gene, thus much faster than SC3.

Usage

get_marker_genes(scExp, y, theta, auc, pvalue, FC, ng, n.cores)

Arguments

scExp	the original expression matrix
y	the clustering results after running SHARP.R
n.cores	number of cores to be used. The default is (n-1) cores, where n is the number of cores in your local computer or server.

Value

a matrix where each row represents each selected marker gene and each column represnts one property of the gene, including its cluster, p-value and AUROC.

Examples

y = SHARP(scExp)
sginfo = get_marker_genes(scExp, y)

shibiaowan/SHARP documentation built on April 28, 2021, 1:56 p.m.