scClusterCells

This tool clusters the cells based on the input count matrix (output of scCountReads) and performs dimentionality reduction, community detection and 2D projection (UMAP) of the cells. The result is an updated loom object, and (optionally) a plot file and a tsv file with UMAP coordinates and corresponding cluster id for each barcode.

usage: Example usage: scClusterCells -i cellCounts.loom -o clustered.loom -op <umap_prefix>.png  > log.txt

Input/Output options

--input, -i

Input file in .loom format

--outFile, -o

The file to write results to. For scFilterStats, scFilterBarcodes and scJSD, the output file is a .txt file. For other tools, the output file is an updated .loom object with the result of the requested operation.

Clustering Options

--outFileUMAP, -op

The output plot file (for UMAP). If you specify this option, another file with the same prefix (and .txt extention) is also created with the raw UMAP coordinates.

--outFileTrainedModel, -om

The output file for the trained LSI model. The saved model can be used later to embed/compare new cells to the existing cluster of cells.

--method, -m

Possible choices: LSA, LDA, glmPCA

The dimentionality reduction method for clustering. (Default: “LSA”)

--glmPCAfamily, -gf

Possible choices: gaussian, poisson, bernoulli, beta, gamma, lognormal, log_normal, sigmoid_beta

The choice of exponential family distribution to use for glmPCA method. (Default: “poisson”)

--binarize

Binarize the counts per region before dimentionality reduction (only for LSA/LDA)

--nPrinComps, -n

Number of principle components to reduce the dimentionality to. Use higher number for samples with more expected heterogenity. (Default: 20)

--nNeighbors, -nk

Number of nearest neighbours to consider for clustering and UMAP. This number should be chosen considering the total number of cells and expected number of clusters. Smaller number will lead to more fragmented clusters. (Default: 30)

--clusterResolution, -cr

Resolution parameter for clustering. Values lower than 1.0 would result in less clusters, while higher values lead to splitting of clusters. In most cases, the optimum value would be between 0.8 and 1.2. (Default: 1.0)

Plot options

--plotWidth

Output plot width (in cm). (Default: 10)

--plotHeight

Output plot height (in cm). (Default: 10)

--plotFileFormat

Possible choices: png, pdf, svg, eps

Image format type. If given, this option overrides the image format based on the plotFile ending. (Default: “png”)

Other options

--verbose, -v

Set to see processing messages.

--version

show program’s version number and exit