eset2Phase: Estimate the phases of expression
In haowulab/SC2P: Two phase differential expression for single-cell RNA-seq

Description Usage Arguments Details Value Author(s) See Also Examples

View source: R/eset2Phase.R

This function estimates the probabilities of expression for all genes in all cells. Given an ExpressionSet, it returns the posterior probability for being in phase II (expression) for each gene in each cell.

1	eset2Phase(eset, low.prob = 0.99)

`eset`	An object of ExpressionSet for the single cell RNA-seq data. This should be constructed using the ExpressionSet function provided by Biobase.
`low.prob`	The probability threshold for calling phase II. When the posterior probability for a gene in a cell is greater than this threshold, it will be called as in phase II (expressed).

The estimation procedure considers the cell-specific (for example, sequencing depth and marginal distribution of expression for all genes), and gene-specific (for example, average expression level for this gene cross all cells) factors, which achieves a data-driven thresholding and provides better results than using an fixed, ad hoc threshold for determining expression state.

An object of sc2pSet, which is an extension of ExpressionSet. Slots in the sc2pSet include the following

`assayData`	This includes three matrices of the same dimension as the original data in the input `eset`. The normalization `Offset` that counts for technical bias such as library size, the estimated probability in Phase 2 (`Z`) and the input counts `exprs`.
`phenoData`	This expanded phenoData includes the same information for each cell(sample) as in `eset`, with additional Phase I parameters for each cell.
`featureData`	This contains the Phase II parameters in the lognormal model for each gene.

annotationInherited from the input eset experimentDataInherited from the input eset

Zhijin (Jean) Wu <zwu@stat.brown.edu>

twoPhaseDE

## Not run: 
data(brain_scRNAseq)

## create ExpressionSet
colnames(Y) <- rownames(design)
phenoData <- new("AnnotatedDataFrame", data=design)
eset <- ExpressionSet(assayData=Y, phenoData=phenoData)

## estimate phases
data <- eset2Phase(eset)
data

## visualize the results for a gene
zyPlot(rownames(data)[1], data, group.name="celltype")

## End(Not run)