neb.train: Nonparametric empirical Bayes classifier without annotations;...
In ssa: Simultaneous Signal Analysis

Description Usage Arguments Value Examples

Treats the control and case minor allele frequencies as random tuples from a bivariate prior distribution G and then estimates the optimal Bayesian classifier given G. Nonparametric maximum likelihood is used as a plug-in estimator for G.

1 2	neb.train(pi0, pi1, n0, n1, d = 25, maxit = 200, tol = 1e-04, verbose = FALSE)

`pi0, pi1`	p x 1 vectors of control and case minor allele frequencies, respectively; IMPORTANT: must be relative to the same allele in both cases and controls
`n0, n1`	number of controls and number of cases, respectively
`d`	if a single number, G is estimated on a d x d grid; if a two-component vector (d0,d1), G is estimated on a d0 x d1 grid
`maxit`	maximum number of EM iterations
`tol`	error tolerance
`verbose`	TRUE to print the error attained by each EM iteration

`Pi0`	grid points for estimating the distribution of the control minor allele frequencies
`Pi1`	grid points for estimating the distribution of the case minor allele frequencies
`D0`	conditional density matrix for controls
`D1`	conditional density matrix for cases
`g`	estimated mixing probability mass function
`P`	proportion of cases

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]));
## 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
neb <- neb.train(colMeans(X0)/2,colMeans(X1)/2,n0,n1,d=c(20,25));
contour(neb$Pi0,neb$Pi1,neb$g);
points(pi0,pi1);