SingleCell10x.jl

A Julia package for loading 10x Single Cell RNA-seq and antibody count data.

It supports loading of 10x ".h5" files as well as text-based ".mtx" and ".tsv/.csv" files. The count matrix can be transposed at load time and the data types of the matrix and annotations can be chosen conveniently.

More information on the file formats can be found in the 10x Genomics offical documentation. Note that this is an independent implementation - any bugs you find should be reported here, not to 10x Genomics.

To analyze your Single Cell expression data in Julia, you can use SingleCellProjections.jl.

Documentation

The package should hopefully be straightforward to use. See below for example usage. More detailed descriptions are in the docstrings of the exported functions:

read10x
read10x_matrix
read10x_features
read10x_barcodes

Example usage

Below, we will load files from the test/ directory, but you can try with your own files instead.

basepath = joinpath(pkgdir(SingleCell10x),"test/data/500_PBMC_3p_LT_Chromium_X_50genes/")

Basic loading

Loading a sparse count matrix, feature annotations and barcode annotations from a ".h5" file:

using SingleCell10x
filepath = joinpath(basepath, "filtered_feature_bc_matrix.h5")

matrix,features,barcodes = read10x(filepath)

You can also load the matrix/features/barcodes separately:

matrix2 = read10x_matrix(filepath)
features2 = read10x_features(filepath)
barcodes2 = read10x_barcodes(filepath)

Sink types

Per default, the matrix will be a SparseMatrixCSC{Int,Int}, features will be a NamedTuple where keys are column names and elements are column vectors, and barcodes a Vector{String}. But you can also specify sink types to load directly into the desired type:

using DataFrames
matrix,features,barcodes = read10x(filepath, Matrix{Float64}, DataFrame, DataFrame)

or similarly

matrix2 = read10x_matrix(filepath, Matrix{Float64})
features2 = read10x_features(filepath, DataFrame)
barcodes2 = read10x_barcodes(filepath, DataFrame)

MatrixMarket .mtx files

Instead of loading a ".h5", SingleCell10x.jl also support the text-based "matrix.mtx(.gz)" with companion files. By default, feature/barcode filenames are searched for in the same folder.

filepath_mtx = joinpath(basepath,"filtered_feature_bc_matrix/matrix.mtx.gz")

matrix,features,barcodes = read10x(filepath_mtx)

If you load features/barcodes separately, you can either specify the path to the "matrix.mtx(.gz)" file or the feature/barcode file directly.

filepath_features = joinpath(basepath,"filtered_feature_bc_matrix/features.tsv.gz")
filepath_barcodes = joinpath(basepath,"filtered_feature_bc_matrix/barcodes.tsv.gz")

features2 = read10x_features(filepath_mtx)
features3 = read10x_features(filepath_features)

barcodes2 = read10x_barcodes(filepath_mtx)
barcodes3 = read10x_barcodes(filepath_barcodes)