predictSmooth: predictSmooth

In statOmics/tradeSeq: trajectory-based differential expression analysis for sequencing data

predictSmooth

Description

Get smoothers estimated by tradeSeq along a grid. This function does not return fitted values but rather the predicted mean smoother, for a user-defined grid of points.

Usage

predictSmooth(models, ...)

## S4 method for signature 'SingleCellExperiment'
predictSmooth(models, gene, nPoints = 100, tidy = TRUE)

## S4 method for signature 'list'
predictSmooth(models, gene, nPoints = 100)

Arguments

models	Either the SingleCellExperiment object obtained after running fitGAM, or the specific GAM model for the corresponding gene, if working with the list output of tradeSeq.
...	parameters including:
gene	Either a vector of gene names or an integer vector, corresponding to the row(s) of the gene(s).
nPoints	The number of points used to create the grid along the smoother for each lineage. Defaults to 100.
tidy	Logical: return tidy output. If TRUE, returns a data.frame specifying lineage, gene, pseudotime and value of estimated smoother. If FALSE, returns matrix of predicted smoother values, where each row is a gene and each column is a point on the uniform grid along the lineage. For example, if the trajectory consists of 2 lineages and nPoints=100, then the returned matrix will have 2*100 columns, where the first 100 correspond to the first lineage and columns 101-200 to the second lineage.

Details

Using the gene expression model of tradeSeq available at https://www.nature.com/articles/s41467-020-14766-3#Sec2. the output of predictSmooth returns the \eta_{gi} value for equally space values of pseudotimes, and a constant value for U_i and N_i (arbitraly, we select the values of i=1).

Value

A matrix with estimated averages.

A vector of fitted values.

Examples

data(gamList, package = "tradeSeq")
predictSmooth(models = gamList, gene = 1)

statOmics/tradeSeq documentation built on Jan. 19, 2024, 8:26 p.m.