findEnhancer: 'findEnhancer' is the main function that identifies enhancers...

In vjbaskar/lenhancer: LASSO based enhancer identification

findEnhancer is the main function that identifies enhancers for a given gene

Description

findEnhancer is the main function that identifies enhancers for a given gene

Usage

findEnhancer(gene, expression, regulation_signal, regulation_tfbs,
  region_gene_mapping, min_tfb_events = 2, coCRE_corr_cutoff = 0.5,
  coCRE_cutoff = 1e+05, singleton_cutoff = 20000, alphaVal = 1,
  scale.predictors = TRUE, family = "gaussian", nfoldxval = 10, ...)

Arguments

gene	gene of interest eg. Runx1 (case sensitive)
expression	data.frame/matrix of expression with cell types as cols and genes as rows
regulation_signal	data.frame/matrix of enhancer signals (active histone marks/DNaseI-seq/ATAC-seq etc) with cell types as cols and putative enhancer regions as rows.
regulation_tfbs	data.frame/matrix with TF ChIP-seq data as cols and putative enhancer regions as rows. This is a binary matrix with 0 = no binding and 1 = binding
region_gene_mapping	data.frame with region to gene mapping. Please refer to data(region_gene_mapping) for the format.
min_tfb_events	the minimum number of TFB events per site to be consider for community CRE calculation. default = 2
coCRE_corr_cutoff	the cutoff above which two putative enhancer regions (CREs) are considered to be correlated. Default = 0.5
coCRE_cutoff	the cutoff beyond which the regions will not be considered for coCRE calculation. Default = 10000 (100kB)
singleton_cutoff	the cutoff for considering singleton CRE. Default = 20000 (20Kb)
alphaVal	alpha value for glmnet. Read glmnet manual
scale.predictors	(TRUE/FALSE). Default = TRUE
family	response family. Default = gaussian
nfoldxval	n-fold cross validation for lmin/l1se. Default = 10 (for leave out cross validation this is equal to total number of cols in expression data)

Value

list of regions and p-values

vjbaskar/lenhancer documentation built on Sept. 22, 2023, 1:29 p.m.