## mfa - Bayesian inference of bifurcations in single-cell data
`mfa` is an R package implementing Gibbs sampling for a Bayesian hierarchichal mixture of factor analysers for inference of bifurcations in single-cell data.
[![Build Status](https://travis-ci.org/kieranrcampbell/mfa.svg?branch=master)](https://travis-ci.org/kieranrcampbell/mfa) [![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.345981.svg)](https://doi.org/10.5281/zenodo.345981)
### Installation
`mfa` may be installed either via Bioconductor (coming soon) or Github:
#### Installation from Bioconductor
```r
if (!requireNamespace("BiocManager", quietly=TRUE))
install.packages("BiocManager")
BiocManager::install("mfa")
```
#### Installation from Github
```r
devtools::install_github("kieranrcampbell/mfa", build_vignettes = TRUE)
```
### Usage
For a cell-by-gene matrix of expression Y, MFA can be envoked via
```r
m <- mfa(Y)
```
which will perform Gibbs sampling to infer pseudotimes, branch structure, and genes involved in the bifurcation.
For full usage see the package vignette:
```r
vignette('introduction_to_mfa')
```
### Authors
Kieran Campbell & Christopher Yau
University of Oxford