ScoreCellsWithSavedModel: Applies a trained binary classifier to get per-cell...

In bimberlabinternal/RIRA_classification: A package for the Rhesus Immune Reference Atlas (RIRA): A multi-tissue single-cell landscape of immune cells

Applies a trained binary classifier to get per-cell probabilities.

Description

Applies a trained model to get per-cell probabilities.

Usage

ScoreCellsWithSavedModel(
  seuratObj,
  model,
  fieldToClass,
  batchSize = 20000,
  assayName = "RNA"
)

Arguments

seuratObj	The Seurat Object to be updated
model	Either the full filepath to a model RDS file, or the name of a built-in model.
fieldToClass	A list mapping the target field name in the seurat object to the classifier level. The latter is either numeric, or the string label. For example: list('CD4_T' = 1, 'CD8_T' = 2))
batchSize	To conserve memory, data will be chunked into batches of at most this many cells
assayName	The assay holding gene expression data

bimberlabinternal/RIRA_classification documentation built on April 14, 2025, 5:59 p.m.