Skip to contents

filter_cells: A wrapper for multiple scRNA filtering methods

filter_cells.Rd

Wrapper function to filter cells from a Seurat object using method of choice. Returns a Seurat object with filtering results

Usage

filter_cells(
  so,
  method = "miQC",
  mads = 3,
  nCount_RNA_max = 50000,
  nCount_RNA_min = 1000,
  nFeature_RNA_max = 5000,
  nFeature_RNA_min = 200,
  percent_mt_max = 10,
  percent_mt_min = 0
)

Arguments

so

Seurat object to filter

method

Filtering method to use: "miQC" (default), "manual", or "mads"

mads

Number of median absolute deviations to use for "mads" method (default 3)

nCount_RNA_max

Maximum nCount_RNA threshold for "manual" method (default 500000)

nCount_RNA_min

Minimum nCount_RNA threshold for "manual" method (default 1000)

nFeature_RNA_max

Maximum nFeature_RNA threshold for "manual" method (default 5000)

nFeature_RNA_min

Minimum nFeature_RNA threshold for "manual" method (default 200)

percent_mt_max

Maximum percent_mt threshold for "manual" method (default 10)

percent_mt_min

Minimum percent_mt threshold for "manual" method (default 0)

Value

Returns a Seurat object with an additional metadata column "keep" indicating whether each cell is to be kept or discarded

Details

This function filters cells from a Seurat object using one of three methods:

  • "miQC": Uses the miQC package to filter cells based on a probabilistic model based on mitochondrialcontent and number of detected features

  • "manual": Filters cells based on user-defined thresholds for nCount_RNA, nFeature_RNA, and percent_mt

  • "mads": Filters cells based on median absolute deviations from the median for nCount_RNA, nFeature_RNA, and percent_mt