clusterPropagation: Batch-wise propagation of clusters through machine learning

In casanova-lab/iMUBAC: Integration of Multi-Batch Cytometry datasets

Batch-wise propagation of clusters through machine learning

Description

clusterPropagation takes both an original SCE object returned by prepSCE and down-sampled, batch-corrected, and clustered SCE object returned by clustering as inputs. This function trains batch-specific classifiers using a specified algorithm, and predicts cluster IDs in a batch-wise manner.

Usage

clusterPropagation(
  sce,
  sce_down,
  by_exprs_values = "exprs",
  maxN = 100,
  numThreads = 4,
  seed = 12345
)

Arguments

sce	A SingleCellExperiment object. colData(sce) must contain file_name and batch columns.
sce_down	A SingleCellExperiment object. colData(sce_down) must contain columns named as file_name, batch, and cluster_id.
by_exprs_values	A character string. Specifies which assay data to use for classifier training and prediction. Must be the non-batch-corrected one in this case.
maxN	Numeric. Specifies the maximum number of unique cells per cluster per batch to be used for classifier training.
numThreads	Numeric. The number of threads for training classifiers.
seed	Numeric. Sets a random seed.

Value

A SingleCellExperiment-class object.

casanova-lab/iMUBAC documentation built on Sept. 13, 2022, 6:36 p.m.