R package for gene signature extraction and cell identity recognition at individual cell level from single-cell RNA-seq.
Welcome to the official Github repository of the CelliD software presented in the Article Gene signature extraction and cell identity recognition at the single-cell level with Cell-ID, Nature Biotechnology 2021
CelliD is a robust statistical method that performs gene signature extraction and functional annotation for each individual cell in a single-cell RNA-seq dataset. CelliD is based on Multiple Correspondence Analysis (MCA) and produces a simultaneous representation of cells and genes in a low dimension space. Genes are then ranked by their distance to each individual cell, providing unbiased per-cell gene signatures. Such signatures proved valuable to (i) correctly predict cell type labels at individual cell resolution, (ii) correctly match cells from the same cell type across independent datasets, overcoming batch effects arising from different technologies, tissues-of-origin and donors, and (iii) uncover functionally relevant cell heterogeneity that would have been missed by clustering-based approaches. CelliD enables the robust identification of rare or even unique cells whose gene signatures are reproducible across diverse single-cell omics datasets.
CelliD has recently moved to Bioconductor but can be installed only with R 4.1. The master branch of this repository is the mirror of the bioconductor package. The build indicates "error" but it is only due to some conflict between Seurat and SingleCellExperiment (as.Seurat, as.SingleCellExperiment function) that should be resolved very soon. The CelliD package is still functionnal.
In order to use CelliD with R version 3.6 ~ 4.0, please install the legacy branch of the repository.
Within R, set first:
install.packages("devtools")
setRepositories(ind = c(1,2,3))To install CelliD then just type:
devtools::install_github("RausellLab/CelliD", ref = "legacy")
library(CellID) # Note that the legacy version is called CellID and not CelliDMAC OS users might experience installation issues related to Gfortran library. To solve such issue download and install the appropriate gfortran dmg file from https://github.com/fxcoudert/gfortran-for-macOS
Installing legacy version with R 3.5/3.6
When installing CelliD from R 3.6 this error might appear.
ERROR: dependency 'Seurat' is not available for package 'CellID'
The Seurat package on CRAN is on version 4 right now and is only usable from R version 4.X.X. install.packages("Seurat") will automatically try to download the version 4.
It is strongly recommended to install R version 4.0 but if you need to install CelliD on R 3.6/3.5 please first install Seurat version 3
remotes::install_version("rsvd", version = "1.0.2")
remotes::install_version("spatstat", version = "1.61.0")
remotes::install_version("Seurat", version = "3.2.3")
And then proceed to install CelliD
setRepositories(ind = c(1,2,3))
devtools::install_github("RausellLab/CelliD", ref = "legacy")
CelliD use as input single cell data in the form of specific S4 objects. Currently supported files are SingleCellExperiment from Bioconductor and Seurat Version 3 or 4 from CRAN.
A vignette illustrating CelliD step-by-step procedures is provided here. Applications include MCA dimensionality reduction, per-cell gene signatures extraction, automatic cell type prediction using marker gene lists, label-transferring across datasets and functional enrichment analysis.
- Akira Cortal - akira.cortal@institutimagine.org
- Antonio Rausell - antonio.rausell@institutimagine.org
This project is licensed under the GNU General Public License 3 - see the LICENSE file for details
Cortal, Akira, Loredana Martignetti, Emmanuelle Six, and Antonio Rausell. “Gene Signature Extraction and Cell Identity Recognition at the Single-Cell Level with Cell-ID.” Nature Biotechnology, April 29, 2021, 1–8. https://doi.org/10.1038/s41587-021-00896-6.
Companion Github repository with R scripts and intermediate data representations required to reproduce all figures from the Cell-ID manuscript can be found here https://github.com/RausellLab/CellIDPaperScript.
You may follow us in Twitter for regular updates: https://twitter.com/AntonioRausell