BFmixture: calculate BF with mixture prior.

In adimitromanolakis/rareBF: Bayesian Models for Rare Variant Association Analysis

Description

calculate BF with mixture prior.

Usage

BFmixture(obs.data, nvariants, low.bound = 10^(-2))

Arguments

obs.data	N*2 data matrix, N is the sample size, 1st column is sum of rare variants for each individual, 2nd column is the disease status
nvariants	number of sites of the region
low.bound	lower bound

Value

Vector of 3 elements. 1st, probability of p=0; 2nd, precision parameter of beta distribution,3rd: BF

Examples

nsites = 500
ncase = 200
ncontrol = 200
maf = 0.005
rexp_param = 2

model1 = function(i) {
 
 p1 = rbinom( ncontrol  , nsites , maf   )
 p2 = rbinom( ncase     , nsites , maf )  + H1 * round(rexp(ncase,rexp_param))
 disease_status = c (  rep(0,ncontrol),rep(1,ncase)  )
 data = data.frame(c(p1,p2), disease_status)
 bf = BFmixture(data, nsites)[3]
 bf
}

nrep = 20

H1 = 0
bf1 = ( replicate(nrep,model1() ) )

H1 = 1
bf2 = ( replicate(nrep,model1() ) )

col = c( rep("blue",nrep) , rep("red", nrep))


plot(log(c(bf1,bf2)),col = col ,pch=20,cex=1.5)

adimitromanolakis/rareBF documentation built on July 24, 2019, 12:26 p.m.