filterCells: Filter Cells

In roryk/bcbioSinglecell: Single-Cell RNA-Seq Utilities

Description

Apply gene detection, novelty score, and mitochondrial abundance cutoffs to cellular barcodes. By default we recommend applying the same filtering cutoff to all samples. The filtering parameters now support per-sample cutoffs, defined using a named numeric vector. When matching per sample, be sure to use the sampleNames() return values (i.e. the sampleName column in sampleData()).

Usage

filterCells(object, ...)

## S4 method for signature 'SingleCellExperiment'
filterCells(object, minUMIs = 0L,
  maxUMIs = Inf, minGenes = 0L, maxGenes = Inf, minNovelty = 0L,
  maxMitoRatio = 1L, minCellsPerGene = 10L)

Arguments

object	Object.
...	Additional arguments.
minUMIs	Minimum number of UMI disambiguated counts per cell.
maxUMIs	Maximum number of UMI disambiguated counts per cell.
minGenes	Minimum number of genes detected.
maxGenes	Maximum number of genes detected.
minNovelty	Minimum novelty score (log10 genes per UMI).
maxMitoRatio	Maximum relative mitochondrial abundance (0-1 scale).
minCellsPerGene	Include genes with non-zero expression in at least this many cells.

Value

bcbioSingleCell, with filtering information slotted into metadata() as filterCells and filterParams.

Author(s)

Michael Steinbaugh

See Also

Seurat::CreateSeuratObject().

Other Quality Control Functions: barcodeRanksPerSample, metrics, plotCellCounts, plotGenesPerCell, plotMitoRatio, plotMitoVsCoding, plotNovelty, plotQC, plotReadsPerCell, plotUMIsPerCell, plotZerosVsDepth

Examples

# SingleCellExperiment ====
object <- cellranger_small
show(object)

x <- filterCells(object, minNovelty = 0L)
show(x)
metadata(x)$filterParams

# Per sample cutoffs
sampleNames(object)
x <- filterCells(
    object = object,
    minUMIs = c(pbmc4k = 100)
)
show(x)
metadata(x)$filterParams

roryk/bcbioSinglecell documentation built on May 27, 2019, 10:44 p.m.