trim_small_groups_and_low_expression_genes: trim_small_groups_and_low_expression_genes
In celaref: Single-cell RNAseq cell cluster labelling by reference

Description Usage Arguments Details Value Examples

View source: R/loading_helper_functions.r

Filter and return a SummarizedExperiment object (dataset_se) by several metrics:

Cells with at least min_lib_size total reads.
Genes expressed in at least min_detected_by_min_samples cells, at a threshold of min_reads_in_sample per cell.
Remove entire groups (clusters) of cells where there are fewer than min_group_membership cells in that group.

1
2
3

trim_small_groups_and_low_expression_genes(dataset_se,
  min_lib_size = 1000, min_group_membership = 5,
  min_reads_in_sample = 1, min_detected_by_min_samples = 5)

`dataset_se`	Summarised experiment object containing count data. Also requires 'ID' and 'group' to be set within the cell information (see `colData()`)
`min_lib_size`	Minimum library size. Cells with fewer than this many reads removed. Default = 1000
`min_group_membership`	Throw out groups/clusters with fewer than this many cells. May change with experiment size. Default = 5
`min_reads_in_sample`	Require this many reads to consider a gene detected in a sample. Default = 1
`min_detected_by_min_samples`	Keep genes detected in this many samples. May change with experiment size. Default = 5

If it hasn't been done already, it is highly reccomended to use this function to filter out genes with no/low total counts (especially in single cell data, there can be many) - without expression they are not useful and may reduce statistical power.

Likewise, very small groups (<5 cells) are unlikely to give useful results with this method. And cells with abnormally small library sizes may not be desireable.

Of course 'reasonable' thresholds for filtering cells/genes are subjective. Defaults are moderately sensible starting points.

A filtered dataset_se, ready for use.

demo_query_se.trimmed  <- 
   trim_small_groups_and_low_expression_genes(demo_query_se)
demo_query_se.trimmed2 <- 
   trim_small_groups_and_low_expression_genes(demo_ref_se, 
                                              min_group_membership = 10)