MAST fits two-part, generalized linear models that are specially adapted for bimodal and/or zero-inflated single cell gene expression data.

MAST supports:

• Easy importing, subsetting and manipulation of expression matrices
• Filtering of low-quality cells
• Adaptive thresholding of background noise
• Tests for univariate differential expression, with adjustment for covariates
• Gene set enrichment analysis, corrected for covariates and gene-gene correlations
• Exploration of gene-gene correlations and co-expression

Vignettes are available in the package via vignette('MAITAnalysis') or vignette('MAST-intro').

• MAST has been ported to use SingleCellExperiment under the hood, and is in Bioconductor.
• We now make an effort to track assay contents (counts vs log counts). This should facilitate interaction with Scater and SCRAN. The older version will remain accessible under branch MASTClassic

For general questions, please submit a question to the bioconductor support site so that others can benefit from the discussion.

For bug reports (something seems broken): open a bug report here.

If you find MAST useful in your work, please consider citing the paper: MAST: a flexible statistical framework for assessing transcriptional changes and characterizing heterogeneity in single-cell RNA sequencing data G Finak, A McDavid, M Yajima, J Deng, V Gersuk, AK Shalek, CK Slichter et al Genome biology 16 (1), 278

If you have previously installed the package SingleCellAssay you will want to remove it as MAST supercedes SingleCellAssay. (If both MAST and SingleCellAssay are attached, opaque S4 dispatch errors will result.) Remove it with:

 remove.packages('SingleCellAssay')

Then you may install or update MAST with:

source("https://bioconductor.org/biocLite.R")
biocLite("MAST")

If you have data analyzed using MASTClassic, you can convert objects from MASTClassic format to the new format based on SingleCellExperiment using convertMastClassicToSingleCellAssay().