cardelino: clone identification from single-cell data

This R package contains a Bayesian method to infer clonal structure for a population of cells using single-cell RNA-seq data (and possibly other data modalities).

Installation

From R

The latest cardelino package can be conveniently installed using the devtools package thus:

devtools::install_github("single-cell-genetics/cardelino", build_vignettes = FALSE)

Cardelino in a container

For situations in which installing software is a challenge (for example, on institutional HPC clusters or on cloud computing platforms), we provide a pre-built Docker image on DockerHub. This image contains R version 3.5.0 with cardelino and other packages (e.g. tidyverse, basic Bioconductor and other single-cell RNA-seq packages) installed and ready to use with Docker or Singularity.

For example, to build a Singularity image that can be used on an HPC cluster (with Singularity installed) one simply pulls the image from DockerHub:

singularity build rsc.img docker://davismcc/r-singlecell-img

This builds a Singularity image called rsc.img in the current working directory. We can then run R from the container and use the installed version of cardelino:

singularity exec rsc.img R

Equivalent commands enable running R from the container with Docker.

Getting started

The best place to start are the vignettes. From inside an R session, load cardelino and then browse the vignettes:

library(cardelino)
browseVignettes("cardelino")

Vignettes for clone identification use cases are provided.

Accessing the vignettes from within your R session is recommended, but you can also view the clone ID vignette.

Notes for donor deconvolution

The denor demultiplex function, namely Vireo, was supported in this R package before, but now has been re-implemented in Python, which is more memory efficient and easier to run via a command line. We, therefore, highly recommend you switch to the Python version: https://vireoSNP.readthedocs.io

The vireo function is not supported from version >=0.5.0. If you want to use the R functions, please use the version ==0.4.2 or lower. You can also find it in a separate branch in this repository: with_vireo branch or use the donor_id.R file directly.

About the name

cardelino is almost an anagram of "clone ID R" and is almost the same as the Italian for "goldfinch", a common and attractive European bird. In the Western art canon, the goldfinch is considered a "saviour" bird and appears in notable paintings from the Italian renaissance and the Dutch Golden Age. Perhaps this package may prove a saviour for certain single-cell datasets.

Vireo(variational inference for reconstructing ensemebl origin) is a Latin word referring to a green migratory bird, perhaps the female golden oriole, possibly the European greenfinch

Acknowledgement: The cardelino image was produced by Darren Bellerby. It was obtained from Flickr and is reproduced here under a CC-BY-2.0 licence.