Skip to content

Latest commit

 

History

History
38 lines (26 loc) · 1.04 KB

README.org

File metadata and controls

38 lines (26 loc) · 1.04 KB

cellTypeCompositions

Estimate cell type compositions from a table of clone by cell type counts. Functions are provided for exploring the cell type compositions that are displayed by cell lineages. The workhorse is a function that performs blocked Gibbs sampling on the cell type frequencies in a parent population under a model that assumes a vague prior for the total number of cells and a uniform Dirichlet prior for the proportions fo different cell types.

Installation

You will need a fairly recent installation of R and the devtools package and its dependencies.

library(devtools)
install_github( "BushmanLab/cellTypeCompositions",
		  ref="main",
		  build_vignettes=TRUE )

Model

The model implemented is described in the model vignette. After installation of the package, enter the following lines at the R prompt:

library(cellTypeCompositions)
vignette( "model", "cellTypeCompositions")

Usage

See the help page for the gibbsScan function. Run the example code for a start.