RunTSNE: Run t-distributed Stochastic Neighbor Embedding

In satijalab/seurat: Tools for Single Cell Genomics

Run t-distributed Stochastic Neighbor Embedding

Description

Run t-SNE dimensionality reduction on selected features. Has the option of running in a reduced dimensional space (i.e. spectral tSNE, recommended), or running based on a set of genes. For details about stored TSNE calculation parameters, see PrintTSNEParams.

Usage

RunTSNE(object, ...)

## S3 method for class 'matrix'
RunTSNE(
  object,
  assay = NULL,
  seed.use = 1,
  tsne.method = "Rtsne",
  dim.embed = 2,
  reduction.key = "tSNE_",
  ...
)

## S3 method for class 'DimReduc'
RunTSNE(
  object,
  cells = NULL,
  dims = 1:5,
  seed.use = 1,
  tsne.method = "Rtsne",
  dim.embed = 2,
  reduction.key = "tSNE_",
  ...
)

## S3 method for class 'dist'
RunTSNE(
  object,
  assay = NULL,
  seed.use = 1,
  tsne.method = "Rtsne",
  dim.embed = 2,
  reduction.key = "tSNE_",
  ...
)

## S3 method for class 'Seurat'
RunTSNE(
  object,
  reduction = "pca",
  cells = NULL,
  dims = 1:5,
  features = NULL,
  seed.use = 1,
  tsne.method = "Rtsne",
  dim.embed = 2,
  distance.matrix = NULL,
  reduction.name = "tsne",
  reduction.key = "tSNE_",
  ...
)

Arguments

object	Seurat object
...	Arguments passed to other methods and to t-SNE call (most commonly used is perplexity)
assay	Name of assay that that t-SNE is being run on
seed.use	Random seed for the t-SNE. If NULL, does not set the seed
tsne.method	Select the method to use to compute the tSNE. Available methods are: “Rtsne”: Use the Rtsne package Barnes-Hut implementation of tSNE (default) “FIt-SNE”: Use the FFT-accelerated Interpolation-based t-SNE. Based on Kluger Lab code found here: https://github.com/KlugerLab/FIt-SNE
dim.embed	The dimensional space of the resulting tSNE embedding (default is 2). For example, set to 3 for a 3d tSNE
reduction.key	dimensional reduction key, specifies the string before the number for the dimension names. “tSNE_” by default
cells	Which cells to analyze (default, all cells)
dims	Which dimensions to use as input features
reduction	Which dimensional reduction (e.g. PCA, ICA) to use for the tSNE. Default is PCA
features	If set, run the tSNE on this subset of features (instead of running on a set of reduced dimensions). Not set (NULL) by default; dims must be NULL to run on features
distance.matrix	If set, runs tSNE on the given distance matrix instead of data matrix (experimental)
reduction.name	dimensional reduction name, specifies the position in the object$dr list. tsne by default

satijalab/seurat documentation built on May 11, 2024, 4:04 a.m.