Skip to contents

ConstructTFNetwork

Source: R/ConstructTFNetwork.R
ConstructTFNetwork.Rd

Construct a network of transcription factors and target genes based on gene co-expression

Usage

ConstructTFNetwork(seurat_obj, model_params, nfold = 5, wgcna_name = NULL)

Arguments

seurat_obj

A Seurat object

model_params

a list of model parameters to pass to xgboost

nfold

number of folds for cross validation

wgcna_name

name of the WGCNA experiment

Value

seurat_obj with the TF network added

Details

ConstructTFNetwork uses motif-gene information to build a directed network of transcription factors (TFs) and target genes. XGBoost regression is leveraged to model the expression of each gene based on its candidate TF regulators. This analysis gives us information about how important each TF is for predicting each gene, allowing us to prioritize the most likely regulators of each gene. This process is done on the gene expression matrix stored with SetDatExpr, which is typically the hdWGCNA metacell gene expression matrix.