cellassign automatically assigns single-cell RNA-seq data to known cell types across thousands of cells accounting for patient and batch specific effects. Information about a priori known markers cell types is provided as input to the model in the form of a (binary) marker gene by cell-type matrix.
cellassign then probabilistically assigns each cell to a cell type, removing subjective biases from typical unsupervised clustering workflows.
Installing from GitHub
cellassign is built using Google's Tensorflow, and as such requires installation of the R package
cellassign can then be installed from github:
install.packages("devtools") # If not already installed devtools::install_github("Irrationone/cellassign")
Installing from conda
With conda, install the current release version of
cellassign as follows:
conda install -c conda-forge -c bioconda r-cellassign
cellassign requires the following inputs:
exprs_obj: Cell-by-gene matrix of raw counts (or SingleCellExperiment with
marker_gene_info: Binary gene-by-celltype marker gene matrix or list relating cell types to marker genes
s: Size factors
X: Design matrix for any patient/batch specific effects
The model can be run as follows:
cas <- cellassign(exprs_obj = gene_expression_data, marker_gene_info = marker_gene_info, s = s, X = X)
An example set of markers for the human tumour microenvironment can be loaded by calling
Please see the package vignette for details and caveats.
Probabilistic cell type assignment of single-cell transcriptomic data reveals spatiotemporal microenvironment dynamics in human cancers, Biorxiv 2019
Allen W Zhang, University of British Columbia
Kieran R Campbell, University of British Columbia