scCombineCounts

scCombineCounts combines multiple count matrices (output of scCountReads) into one, either assuming they are different samples (multi-sample) or different measurements on the same set of cells (multi-modal). The result is a .h5ad (AnnData) file with combined counts in multi-sample mode or a .h5mu (MuData) file in multi-modal mode. NOTE: it doesn't perform any 'batch effect correction' or 'integration' of data from different technologies, which requires more sophisticated methods.

usage: 
scCombineCounts -i sample1.h5ad sample2.h5ad -o combined.h5ad -m multi-sample
scCombineCounts -i modality1.h5ad modality2.h5ad -o combined.h5mu -m multi-modal

Input/Output options

--input, -i

List of .h5ad files separated by spaces.

General Options

--outFile, -o

The file to write results to. For method: multi-sample, the output file is an updated .h5ad object, which can be used by other tools. For method: multi-modal, the output file is an .h5mu file. This file can only be used by scClusterCells, to perform multi-modal clustering.

--labels, -l

User defined labels instead of default labels from file names. Multiple labels have to be separated by a space, e.g. --labels sample1 sample2 sample3. In case of --method 'multi-modal' the --labels can be used as the labels of different data modalities (eg. RNA, ATAC)

--method, -m

Possible choices: multi-sample, multi-modal

How to merge the counts from the provided samples. multi-sample: assumes that each sample is independent, but were counted in the same manner (i.e. on same features), therefore it looks for feature overlaps, but not for barcode overlaps. multi-modal: assumes that the counts were generated in 2 different ways, but from the same set of cells (for example, using a multi-omic assay), therefore it looks for the overlap of cell barcodes, but not for the overlaps of features (Default: 'multi-sample')

Other options

--verbose, -v

Set to see processing messages.

--version

show program's version number and exit