In drisso/scRNAseq: Collection of Public Single-Cell RNA-Seq Datasets

library(BiocStyle)
knitr::opts_chunk$set(error=FALSE, message=FALSE, warning=FALSE)

Downloading the count data

We obtain a single-cell RNA sequencing dataset of Xenopus tail cells from @aztekin2019identification. We download and cache the count matrix and metadata using the r Biocpkg("BiocFileCache") package.

library(BiocFileCache)
bfc <- BiocFileCache("raw_data", ask = FALSE)
contents <- bfcrpath(bfc,
    file.path("https://www.ebi.ac.uk/arrayexpress/files",
        "E-MTAB-7716/E-MTAB-7716.processed.1.zip"))

unzip(contents, exdir=tempdir())
unzip(file.path(tempdir(), "arrayExpressUpload.zip"), exdir=tempdir())
unzip(file.path(tempdir(), "ArrayExpressV2.zip"), exdir=tempdir())

Processing the read counts

We load the counts into memory. Despite the name of the file, it is not actually comma-separated!

library(Matrix)
path <- file.path(tempdir(), "ArrayExpress")
counts <- readMM(file.path(path, "countsMatrix.mtx"))
counts <- as(counts, "dgCMatrix")
rownames(counts) <- read.csv(file.path(path, "genes.csv"), 
    stringsAsFactors=FALSE, header=FALSE)[,1]
colnames(counts) <- read.csv(file.path(path, "cells.csv"), 
    stringsAsFactors=FALSE, header=FALSE)[,1]
dim(counts)

Processing the metadata

Processing the per-cell metadata, and merging in the per-sample metadata along with it.

library(S4Vectors)
meta <- read.csv(file.path(path, "meta.csv"), stringsAsFactors=FALSE)
labels <- read.csv(file.path(path, "labels.csv"), stringsAsFactors=FALSE)
meta <- cbind(meta, labels[match(meta$sample, labels$Sample),-1])
meta <- DataFrame(meta)
rownames(meta) <- NULL
meta

Now enforcing consistency checks with the column names of counts.

stopifnot(identical(colnames(counts), meta$cell))

Saving for upload

We save these to file for upload to r Biocpkg("ExperimentHub").

repath <- file.path("scRNAseq", "aztekin-tail", "2.0.0")
dir.create(repath, showWarnings=FALSE, recursive=TRUE)
saveRDS(counts, file=file.path(repath, "counts.rds"))
saveRDS(meta, file=file.path(repath, "coldata.rds"))