scSEGIndex: Single Cell Stably Express Gene Index

In SydneyBioX/scMerge: scMerge: Merging multiple batches of scRNA-seq data

Single Cell Stably Express Gene Index

Description

This function computes the single-cell Stably Expressed Gene (scSEG) index from Lin. et al. (2019) for a given single-cell count data matrix. Each gene in the data is fitted with a gamma-normal mixture model and the final SEG index is computed as an average of key parameters that measure the expression stability of a gene.

We recommend using either the pre-computed genes (see "See Also" below) or the top SEG genes from an user's own data as the control genes in the scMerge function (see the ctl argument in the scMerge function).

Usage

scSEGIndex(
  exprs_mat,
  cell_type = NULL,
  BPPARAM = SerialParam(progressbar = TRUE),
  return_all = FALSE
)

Arguments

exprs_mat	A log-transformed single-cell data, assumed to have no batch effect and covered a wide range of cell types. A n by m matrix, where n is the number of genes and m is the number of cells.
cell_type	A vector indicating the cell type information for each cell in the gene expression matrix. If it is NULL, the function calculates the scSEG index without using F-statistics.
BPPARAM	A BiocParallelParam class object from the BiocParallel package is used. Default is SerialParam(progressbar = TRUE).
return_all	Default to FALSE. If set to TRUE, then all genes will be returned. Otherwise, only genes with less than 80 percent zeroes will be returned.

Value

Returns a data frame. Each row is a gene and each column is a statistic relating to the stability of expression of each gene. The main statistic is the segIdx column, which is the SEG index.

Author(s)

Shila Ghazanfar, Yingxin Lin, Pengyi Yang

References

Evaluating stably expressed genes in single cells (2019). doi:10.1093/gigascience/giz106.

See Also

Download human SEG directly from this link; Download mouse SEG directly from this link.

Examples

## Loading example data
data('example_sce', package = 'scMerge')
## subsetting genes to illustrate usage.
exprs_mat = SummarizedExperiment::assay(example_sce, 'logcounts')[1:110, 1:20]
set.seed(1)
result1 = scSEGIndex(exprs_mat = exprs_mat)
## If parallelisation is needed:
param = BiocParallel::MulticoreParam(workers = 2, progressbar = TRUE)
result2 = scSEGIndex(exprs_mat = exprs_mat, BPPARAM = param)
## Closing the parallelisation
BiocParallel::register(BPPARAM = BiocParallel::SerialParam())

SydneyBioX/scMerge documentation built on April 14, 2024, 3:22 a.m.