SimpleIntegration: Integrate Seurat objects simply and easily

In j-andrews7/EZscRNA: Simple Convenience Functions for High-Level scRNA-Seq Analysis

Description

SimpleIntegration integrates Seurat objects using either Seurat integration functions or RunFastMNN, a wrapper from SeuratWrappers around fastMNN from batchelor. This is useful for removing batch effects. It can take either a list of Seurat objects or a meta.data variable to split a single Seurat object by prior to integration.

Usage

SimpleIntegration(scrnas, split.by = NULL, method = c("Seurat", "MNN"),
  vars.to.regress = NULL, n.features = 3000)

Arguments

scrnas	Either a list of Seurat objects or a single Seurat object, the latter of which requires split.by to be provided as well.
split.by	String containing name of meta.data column to be used to split Seurat object into multiple samples.
method	String indicating method to be used. "Seurat" will use Seurat's internal functions for SCT integration. "MNN" will use RunFastMNN.
vars.to.regress	Vector of meta.data variables to be regressed out during SCTransform.
n.features	Number of anchor features to use with SelectIntegrationFeatures.

Details

Performs SCTransform on each Seurat object individually prior to integration.

If method="MNN" is used, SeuratWrappers must be installed.

Value

An integrated Seurat object with technical variation/batch effects removed from the individual samples.

Author(s)

Jared Andrews

Examples

## Not run: 
library(Seurat)
# Will error due to too few cells.
pbmc.int <- SimpleIntegration(pbmc_small, split.by = "groups")

## End(Not run)

j-andrews7/EZscRNA documentation built on Feb. 24, 2020, 10:37 a.m.