echinus

Get a gene expression table from counts for equivalence classes (ECs -> GE table)

Echinus provides a fast & easy way to summarise transcript compatibility counts to gene abundances. Counts for equivalence classes corresponding to one gene are summed up in order to create a cells x genes matrix commonly used for downstream analysis.

Please address documentation for kallisto and bustools for details on getting the TCCs for your data. In particular, check out the possibility to use kallisto bus command to pseudo-align reads while handling your barcode-UMI design via -x argument.

Usage

Echinus can be used on TCCs that are produced by kallisto pseudo command:

echinus count --tcc matrix.tcc.mtx --ecmap matrix.ec --txnames transcripts.txt --genemap transcripts_to_genes.txt --output echinus_dir 

or on the output of the kallisto|bustools pipeline after correcting the barcodes and sorting the BUS file:

echinus count --bus output.correct.sort.bus --ecmap matrix.ec --txnames transcripts.txt --genemap transcripts_to_genes.txt --output echinus_dir --cells barcodes_whitelist.txt

The output of echinus is represented in 3 files saved to the directory provided:

1. matrix.mtx containing a sparse matrix of gene counts (barcodes in rows),

2. barcodes.tsv containing a list of barcodes,

3. genes.tsv with tab-separated gene IDs and gene names.

The matrix can be easily loaded into R with readMM, or with Read10X when using Seurat for downstream analysis, and into Python with scipy.io.mmread or scanpy.api.read_10x_mtx.

Implementation details

Echinus uses cobra's implementation of CLI.