scMEB: scMEB: A fast and clustering-independent method for detecting...

In FocusPaka/MEB: A normalization-invariant minimum enclosing ball method to detect differentially expressed genes for RNA-seq data

scMEB: A fast and clustering-independent method for detecting differentially expressed genes in single-cell RNA-seq data

Description

Using the Minimum Enclosing Ball (MEB) method to discriminate differential expression genes (DEGs) without prior cell clustering results.

Usage

scMEB(sce, stable_idx, filtered = FALSE, 
gamma = seq(1e-04,0.001,1e-05), nu = 0.01, reject_rate = 0.1)

Arguments

sce	A SingleCellExperiment class scRNA-seq data.
stable_idx	A vector shows the name of stable expressed gene in sce.
filtered	A logical value to show if the data have been filtered.
gamma	A parameter needed for all kernels except linear.
nu	A parameter needed for one-classification.
reject_rate	A value used in controling the scale of ball, default is 0.01.

Value

list(.) A list of results, "model" represents the model of scMEB, which could be used to discriminate a new gene, "dat_pca" represents the first 50 PCs of each genes, "gamma" represents the selected gamma parameters in model scMEB, "train_error" represents the corresponding train_error when the value of gamma changed, "dist" shows the distance between the points and the radius of the sphere in feature space.

Examples

## Simulation data for scRNA-seq data generated from splatter package.
library(SingleCellExperiment)
data(sim_scRNA_data)
data(stable_gene)
sim_scRNA <- scMEB(sce=sim_scRNA_data, stable_idx=stable_gene, 
filtered = FALSE, gamma = seq(1e-04,0.001,1e-05), nu = 0.01, 
reject_rate = 0.1)

FocusPaka/MEB documentation built on April 23, 2023, 5:40 p.m.