nebula.chisq.predict: Nonparametric empirical Bayes classifier using latent...

In ssa: Simultaneous Signal Analysis

Description

Nonparametric empirical Bayes classifier using latent annotations: chi-square test statistics; prediction

Usage

nebula.chisq.predict(newX, nebula, P = NULL, cores = 1)

Arguments

newX	n x p matrix of additively coded genotypes to be predicted; IMPORTANT: must be coded relative to the same allele as in the cases and controls
nebula	output of nebula.chisq.train()
P	prevalence of cases in the testing set; if NULL, P is taken from the train object
cores	number of cores to use

Value

ll	2 x p matrix of log-likelihoods, first row is from controls
score	risk score
class	predicted class, 0=control, 1=case

Examples

p <- 1000; ## number of snps
I <- rep(0,p); I[1:10] <- 1; ## which snps are causal
set.seed(1); pi0 <- runif(p,0.1,0.5); ## control minor allele frequencies
set.seed(1); ors <- runif(sum(I),-1,1); ## odds ratios
pi1 <- pi0;
pi1[I==1] <- expit(ors+logit(pi0[I==1]));
set.seed(1); lam <- rep(0,p); lam[I==1] <- rchisq(sum(I==1),1,50); ## ncps
## training data
n0 <- 100; ## number of controls
X0 <- t(replicate(n0,rbinom(p,2,pi0))); ## controls
n1 <- 50; ## number of cases
X1 <- t(replicate(n1,rbinom(p,2,pi1))); ## cases
T <- rchisq(p,1,lam); ## chi-square statistics
nebula <- nebula.chisq.train(colMeans(X0)/2,colMeans(X1)/2,n0,n1,T,d=c(10,12,14));
## testing data
newX <- rbind(t(replicate(n0,rbinom(p,2,pi0))),
              t(replicate(n1,rbinom(p,2,pi1))));
newY <- c(rep(0,n0),rep(1,n1));
Yhat <- nebula.chisq.predict(newX,nebula);
mean(abs(newY-Yhat$class));

ssa documentation built on May 1, 2019, 10:27 p.m.