README.md

CellID: gene signature extraction and cell identity recognition at individual cell level

CellID take as input a matrix of expression level generated by SCRNA-Seq and performs dimensionality reduction using Multiple Corespondance Analysis. This enables a representation of both genes and cell in the same euclidean space with genes close to a particular cell or group of cells being specificly expressed in those. This property allows to get a ranking of specificity of genes for each cells giving the opportunity to extract meaningful geneset or perform geneset enrichment analysis without any need of clustering.

Installation

CellID contains dependencies with Biocondutor packages. Please use first setRepositories() and type 1 2, to enable the download of bioconducor package.

install.packages("devtools")
library(devtools)
install_github("cbl-imagine/CellID")

Alternatively you can download all bioconductor dependencies beforehand.

if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("scater", version = "3.8")
BiocManager::install("fgsea", version = "3.8")

python umap-learn package can be handy for fast umap computation but is not mandatory.

pip install umap-learn

Usage

CellID is used with single cell specific S4objects. Curreltly supported are SingleCellExperiment from Bioconductor and Seurat Version 3 from CRAN.

Vignettes

You can find a vignette for the basic functionalities and usage of the package here

Authors

License

This project is licensed under the GNU General Public License - see the LICENSE file for details

References

Acknowledgments

A big thanks to the fast growing SingleCell community that keeps giving and delivering amazing data and software.



cbl-imagine/cellID documentation built on Nov. 12, 2019, 1:36 a.m.