enhancerToGene: enhancerToGene

In aertslab/ScoMAP: ScoMAP: Single-Cell Omics Mapping into spatial Axes using Pseudotemporal ordering

Description

Link enhancers to target genes

Usage

enhancerToGene(
  VM_RNA_mat,
  VM_ATAC_mat,
  searchSpace,
  method,
  minOverlap = 0.4,
  nCores = 1,
  nTrees = 1000
)

Arguments

VM_RNA_mat	Matrix containing regions as rows, virtual cells as columns and cisTopic's region accessibility probabilities as values (recommended, see getVirtualFeatureMatrix())
searchSpace	Search space GenomicRanges object as obtained from getSearchSpace()
method	Whether to use pearson correlation between region accessibility and gene expression ('Correlation') or a random forest ('RF') model to infer enhancer-to-gene links.
nCores	How many cores to use if method='RF'. Default: 1
nTrees	How many trees to use per RF model if method='RF'. Default: 1000
minoverlap	Minimum overlap between the candidate regulatory regions and the search space. Default: 0.4.

Value

A list containing a data frame for each gene, in which the RF importance (if method='RF') or the correlation values (if method='Correlation') are given.

Examples

RF_links <- enhancerToGene(VM_DGEM, VM_PRTM, searchSpace, method='RF') 
Cor_links <- enhancerToGene(VM_DGEM, VM_PRTM, searchSpace, method='Correlation') 

aertslab/ScoMAP documentation built on Oct. 8, 2021, 1:15 a.m.