tSNEplot: t-Distributed Stochastic Neighbor embedding with tSNEplot

In systemPipeR/spsBio: spsBio: A systemPipeShiny plugin for biological data visualization

Description

This function computes and plots t-Distributed Stochastic Neighbor embedding (t-SNE) analysis for unsupervised nonlinear dimensionality reduction of count expression matrix. Internally, it is applied the Rtsne::Rtsne() function, using the exact t-SNE computing with theta=0.0.

Usage

tSNEplot(
  countMatrix,
  targets,
  plotly = FALSE,
  savePlot = FALSE,
  filePlot = NULL,
  ...
)

Arguments

countMatrix	date.frame or matrix containing raw read counts.
targets	targets data.frame.
plotly	logical: when FALSE (default), the ggplot2 plot will be returned. TRUE option returns the plotly version of the plot.
savePlot	logical: when FALSE (default), the plot will not be saved. If TRUE the plot will be saved, and requires the filePlot argument.
filePlot	file name where the plot will be saved. For more information, please consult the ggplot2::ggsave() function.
...	additional parameters for the Rtsne::Rtsne() function.

Value

returns an object of ggplot or plotly class.

Examples

targetspath <- system.file("extdata", "targets.txt", package="systemPipeR")
targets <- read.delim(targetspath, comment="#")
cmp <- systemPipeR::readComp(file=targetspath, format="matrix", delim="-")
countMatrixPath <- system.file("extdata", "countDFeByg.xls", package="systemPipeR")
countMatrix <- read.delim(countMatrixPath, row.names=1)
set.seed(42) ## Set a seed if you want reproducible results
tSNEplot(countMatrix, targets, perplexity = 5)

systemPipeR/spsBio documentation built on Oct. 2, 2020, 9:30 a.m.