scFeatures: Wrapper function to run all feature types in scFeatures

In SydneyBioX/scFeatures: scFeatures: Multi-view representations of single-cell and spatial data for disease outcome prediction

Wrapper function to run all feature types in scFeatures

Description

The scFeatures function generates a variety of features from a Seurat object containing single cell RNA-sequencing data. By default, all feature types will be generated and returned in a single list containing multiple data frames.

Usage

scFeatures(
  data = NULL,
  sample = NULL,
  celltype = NULL,
  spatialCoords = NULL,
  spotProbability = NULL,
  feature_types = NULL,
  type = "scrna",
  ncores = 1,
  species = "Homo sapiens",
  celltype_genes = NULL,
  aggregated_genes = NULL,
  geneset = NULL
)

Arguments

data	input data, a matrix of genes by cells
sample	a vector of sample information
celltype	a vector of cell type information
spatialCoords	a list of two vectors containing the x and y coordinates of each cell
spotProbability	a matrix of spot probability, each row represents a celltype and each column represents a spot
feature_types	vector containing the name of the feature types to generate, options are "proportion_raw", "proportion_logit" , "proportion_ratio", "gene_mean_celltype", "gene_prop_celltype", "gene_cor_celltype", "pathway_gsva" , "pathway_mean", "pathway_prop", "CCI", "gene_mean_aggregated", "gene_prop_aggregated", 'gene_cor_aggregated', "L_stats" , "celltype_interaction" , "morans_I", "nn_correlation". If no value is provided, all the above feature types will be generated.
type	input data type, either "scrna" (stands for single-cell RNA-sequencing data), "spatial_p" (stands for spatial proteomics data), or "spatial_t" (stands for single cell spatial data )
ncores	number of cores , default to 1
species	either "Homo sapiens" or "Mus musculus". Defaults to "Homo sapiens" if no value provided
celltype_genes	the genes of interest for celltype specific gene expression feature category If no value is provided, the top variable genes will be used
aggregated_genes	the genes of interest for overall aggregated gene expression feature category If no value is provided, the top variable genes will be used
geneset	the geneset of interest for celltype specific pathway feature category If no value is provided, the 50 hallmark pathways will be used

Value

a list of dataframes containing the generated feature matrix in the form of sample x features

Examples

utils::data("example_scrnaseq" , package = "scFeatures") 
data <- example_scrnaseq
celltype <- data$celltype
sample <- data$sample
data <- data@assays$RNA@data
scfeatures_result <- scFeatures(data, celltype = celltype, sample = sample, type = "scrna", feature_types = "proportion_raw")

SydneyBioX/scFeatures documentation built on Aug. 29, 2024, 10:37 a.m.