Predict_doublets: Predict doublets

In rbr1/MLtiplet_2.0: scRNA-seq doublet/multiplet detection using multi-omic profiling

Predict doublets

Description

This function builds classifier based on "variate_names" plus "marker_array' (module scores for each cell type), and runs the classifier over the whole dataset. This assumes that the doublets_training are all cell IDs wihtin the seurat object.

Usage

Predict_doublets(object, marker_array, doublets_training, variables_norm,
  training_trim = 1)

Arguments

object	= seurat object
marker_array	= a matrix of the module scores per cell types/cluster per cell.
doublets_training	= an array of the training doublet/multiplet set (e.g. from VDJ-seq, CITE-seq or other).
variables_norm	= a matrix of the normalised variables per cell.
training_trim	= a parameter for filtering outliers from the training doublet/multiplet set (default = 1).

Value

doublet object (list of 1. the training doublet prediction table, 2. the doublet probabilities excluding the

training doublets, 3. the doublet probabilities including the training doublets, 4. the list of predicted doublets).

rbr1/MLtiplet_2.0 documentation built on April 11, 2024, 7:02 p.m.