extractCounts: Function for filtering count data

In RaceID: Identification of Cell Types, Inference of Lineage Trees, and Prediction of Noise Dynamics from Single-Cell RNA-Seq Data

Function for filtering count data

Description

This function discards lowly expressed genes from a count matrix stored in an SCseq object, and returns (normalized or non-normalized) gene expression or noise values.

Usage

extractCounts(
  object,
  minexpr = 5,
  minnumber = 5,
  noise = FALSE,
  pt = NULL,
  n = NULL,
  g = NULL,
  norm = TRUE
)

Arguments

object	SCseq class object.
minexpr	Integer number greater or equal to zero. Minimum expression of a gene in at least minnumber cells to not be discarded. Default is 5.
minnumber	Integer number greater or equal to zero. Minimum number of cells required to have at least minexpr transcript counts for a gene to not be discarded. Default is 5.
noise	logical. If TRUE, then noise (in object@noise) is returned for the filtered genes and cells. Default is FALSE and gene expression counts are returned.
pt	List object returned by function pseudoTime. If given, then feature matrix is returned for cells in pt$ord and ordered by pseudo-time. Default is NULL and feature matrix is returned for all cells in object$ndata.
n	Vector of valid column names corresponding to a subset of valid column names of the object@ndata. Default is NULL filtering is done on all cells in object@ndata. Only considered if pt is NULL.
g	Vector of gene IDs (valid row names of object@ndata). If given, then all genes not in g are discarded prior to filtering. Default is NULL and filtering is done on all genes in object@ndata.
norm	logical. If TRUE, then transcipt counts are normalized to the minimum number of total transcript counts across all cells in the feature matrix.

Value

Filtered expression matrix.

RaceID documentation built on Sept. 28, 2023, 5:06 p.m.