RunSPCA: Run Supervised Principal Component Analysis

View source: R/generics.R

RunSPCAR Documentation

Run Supervised Principal Component Analysis

Description

Run a supervised PCA (SPCA) dimensionality reduction supervised by a cell-cell kernel. SPCA is used to capture a linear transformation which maximizes its dependency to the given cell-cell kernel. We use SNN graph as the kernel to supervise the linear matrix factorization.

Usage

RunSPCA(object, ...)

## Default S3 method:
RunSPCA(
  object,
  assay = NULL,
  npcs = 50,
  reduction.key = "SPC_",
  graph = NULL,
  verbose = FALSE,
  seed.use = 42,
  ...
)

## S3 method for class 'Assay'
RunSPCA(
  object,
  assay = NULL,
  features = NULL,
  npcs = 50,
  reduction.key = "SPC_",
  graph = NULL,
  verbose = TRUE,
  seed.use = 42,
  ...
)

## S3 method for class 'Assay5'
RunSPCA(
  object,
  assay = NULL,
  features = NULL,
  npcs = 50,
  reduction.key = "SPC_",
  graph = NULL,
  verbose = TRUE,
  seed.use = 42,
  layer = "scale.data",
  ...
)

## S3 method for class 'Seurat'
RunSPCA(
  object,
  assay = NULL,
  features = NULL,
  npcs = 50,
  reduction.name = "spca",
  reduction.key = "SPC_",
  graph = NULL,
  verbose = TRUE,
  seed.use = 42,
  ...
)

Arguments

object

An object

...

Arguments passed to other methods and IRLBA

assay

Name of Assay SPCA is being run on

npcs

Total Number of SPCs to compute and store (50 by default)

reduction.key

dimensional reduction key, specifies the string before the number for the dimension names. SPC by default

graph

Graph used supervised by SPCA

verbose

Print the top genes associated with high/low loadings for the SPCs

seed.use

Set a random seed. By default, sets the seed to 42. Setting NULL will not set a seed.

features

Features to compute SPCA on. If features=NULL, SPCA will be run using the variable features for the Assay.

layer

Layer to run SPCA on

reduction.name

dimensional reduction name, spca by default

Value

Returns Seurat object with the SPCA calculation stored in the reductions slot

References

Barshan E, Ghodsi A, Azimifar Z, Jahromi MZ. Supervised principal component analysis: Visualization, classification and regression on subspaces and submanifolds. Pattern Recognition. 2011 Jul 1;44(7):1357-71. https://www.sciencedirect.com/science/article/pii/S0031320310005819?casa_token=AZMFg5OtPnAAAAAA:_Udu7GJ7G2ed1-XSmr-3IGSISUwcHfMpNtCj-qacXH5SBC4nwzVid36GXI3r8XG8dK5WOQui;


Seurat documentation built on May 29, 2024, 4:20 a.m.