computeParamPvals: Computes per parameter p-value

In ndukler/superEnhancerModelR: superEnhancerModelR

Description

Computes a p-value for the intercept and each enhancer coefficient in the link function by performing likelihood ratio tests between the full model and a reduced model without that coefficient. Can be parallelized if doParallel is installed. NOTE. Make sure your full model is already fit before running this function!!!

Arguments

x	enhancerDataObject
threads	number of threads to use for refiting, requires parallel
...	all aditional parameters to be passed to optimizer function optimDE

Value

A list with two elements:

• pVals: A table of the coefficient p-values

• reducedModels: Refitted reduced models used to estimate the p-values

Examples

## Create a test design data.frame
design=expand.grid(E1=c(0,1),E2=c(0,1),E3=c(0,1))
## Create fake expression data
expression=c(0.001,0.2,0.3,0.7,0.4,0.8,0.6,1)*100
## Create activity function
actFun=formula(~E1+E2+E3+E1:E2)
edo = enhancerDataObject(expression,design,actFun)
edo=optimDE(edo,maxit=500)
param.stats=computeParamPvals(edo,maxit=500)

ndukler/superEnhancerModelR documentation built on May 17, 2019, 8:18 p.m.