Skip to contents

Filter genes based on the number of expressing cells

Source: R/preprocessing.R
sn_filter_genes.Rd

This function filters genes in a Seurat object based on the number of cells in which they are expressed. Optionally, it can visualize the effect of different filtering thresholds and retain only genes matching bundled GENCODE-based annotation classes.

Usage

sn_filter_genes(
  x,
  min_cells = 3,
  plot = TRUE,
  filter = TRUE,
  assay = "RNA",
  layer = "counts",
  species = NULL,
  gene_class = NULL,
  gene_type = NULL
)

Arguments

x

A Seurat object.

min_cells

An integer specifying the minimum number of cells in which a gene must be expressed to be retained. Default is 3.

plot

Logical; if TRUE, a bar plot is generated showing the number of remaining genes at different filtering thresholds. Default is TRUE.

filter

Logical; if TRUE, returns the filtered Seurat object. If FALSE, returns the original object. Default is TRUE.

assay

Assay used when extracting expression values. Defaults to "RNA".

layer

Layer used for gene filtering. Defaults to "counts".

species

Optional species label used for annotation-aware filtering. Required only when it cannot be inferred from the object.

gene_class

Optional coarse annotation class to retain. One of "coding" or "noncoding".

gene_type

Optional character vector of exact GENCODE gene_type values to retain.

Value

A filtered Seurat object if filter = TRUE, otherwise the original object.

Details

The function computes the number of cells expressing each gene in the Seurat object and filters out genes expressed in fewer than min_cells cells. If plot = TRUE, it visualizes the effect of filtering using ggplot2. When gene_class or gene_type is supplied, the function also matches the feature set against the bundled shennong_gene_annotations data and keeps only the requested annotation subset.

Examples

if (requireNamespace("Seurat", quietly = TRUE)) {
  data("pbmc_small", package = "Shennong")
  pbmc_filtered <- sn_filter_genes(
    pbmc_small,
    min_cells = 5,
    plot = FALSE,
    filter = TRUE
  )
  pbmc_coding <- sn_filter_genes(
    pbmc_small,
    min_cells = 1,
    plot = FALSE,
    filter = TRUE,
    species = "human",
    gene_class = "coding"
  )
}
#> WARN [2026-03-26 18:52:18] Annotation-based gene filtering could not match 17 features for species 'human'. Those unmatched features will be dropped.