tooManyCells {TooManyCellsR}

Execute 'too-many-cells'.

Description

This function will run 'too-many-cells' on a Matrix. Requires 'TooManyCells' to be installed (follow instructions at https://gregoryschwartz.github.io/too-many-cells/ ).

Usage

tooManyCells(mat, args = c("make-tree"), labels = NULL,
  output = "out", prior = NULL, docker = NULL, mounts = c())

Arguments

mat

The input Matrix with gene row names and cell barcode column names.

args

The arguments to give to the command line program. See https://gregoryschwartz.github.io/too-many-cells/ for more information. Defaults to "make-tree".

labels

The input labels data frame with item (cell barcodes) and label (whatever labels you want to give them, such as tissue of origin, celltype, etc.) columns. Optional.

output

The output folder for the 'too-many-cells' process. Defaults to "out".

prior

The location of the tree that was already made (previous 'too-many-cells' output) so quick visual or pruning changes can be made without remaking the tree (can potentially save hours).

docker

If using 'too-many-cells' with docker, use this argument as the command to call. For instance, if version 0.2.1.0 was pulled from Docker Hub, set to "gregoryschwartz/too-many-cells:0.2.1.0".

mounts

Additional directories to mount if needed for docker. The 'prior' argument will automatically mount if specified.

Value

A list of each output, including the stdout. Reads the following files, see https://gregoryschwartz.github.io/too-many-cells/ for more details: "dendrogram.svg", "clumpiness.pdf", "projection.pdf", "label_projection.pdf", "clumpiness.csv", "cluster_info.csv", "node_info.csv", and "cluster_diversity.csv".

Examples

input <- system.file("extdata", "mat.csv", package="TooManyCellsR")
inputLabels <- system.file("extdata", "labels.csv", package="TooManyCellsR")
df = read.csv(input, row.names = 1, header = TRUE)
mat = Matrix::Matrix(as.matrix(df), sparse = TRUE)
labelsDf = read.csv(inputLabels, header = TRUE)
# Here we draw this small toy example with no filter or normalization, and
# decrease the size of the branches and increase the size of the leaf nodes.
# With non-toy real world single cell data, these options should not be
# necessary.
## Not run: 
res = tooManyCells( mat, labels = labelsDf
                  , args = c( "make-tree"
                            , "--no-filter"
                            , "--normalization", "NoneNorm"
                            , "--draw-max-node-size", "40"
                            , "--draw-max-leaf-node-size", "70"
                            )
                  )
plot(res$treePlot, axes = FALSE)
res$stdout
res$nodeInfo

## End(Not run)