DIME (version 1.3.0)

gng.classify: Classification Based on GNG Model

Description

Classifies observed data into differential and non-differential groups based on GNG model.

Usage

gng.classify(data, obj, obj.cutoff = 0.1, obj.sigma.diff.cutoff = NULL, 
  obj.mu.diff.cutoff = NULL)

Arguments

data

an R list of vector of normalized intensities (counts). Each element can correspond to particular chromosome. User can construct their own list containing only the chromosome(s) they want to analyze.

obj

a list object returned by gng.fit function.

obj.cutoff

optional local fdr cutoff for classifying data into differential and non-differential groups based on GNG model.

obj.sigma.diff.cutoff

optional cut-off for standard deviation of the normal component in the best model to be declared as representing differential.

obj.mu.diff.cutoff

optional cut-off for standard deviation of the normal component in the best model to be declared as representing differential.

Value

A list object passed as input with additional element $class containing vector of classifications for all the observations in data. A classification of 1 denotes that the data is classified as differential with fdr < obj.cutoff.

mu.diff.cutoff

normal component with mean > mu.diff.cutoff will be used to represent differential component.

sigma.diff.cutoff

normal component with standard deviation > sigma.diff.cutoff will be used to represent differential component.

See Also

gng.fit

Examples

# NOT RUN {
library(DIME);
# generate simulated datasets with underlying exponential-normal components
N1 <- 1500; N2 <- 500; K <- 4; rmu <- c(-2.25,1.50); rsigma <- c(1,1); 
rpi <- c(.05,.45,.45,.05); rbeta <- c(12,10);
set.seed(1234);
chr1 <- c(-rgamma(ceiling(rpi[1]*N1),shape = 1,scale = rbeta[1]), 
  rnorm(ceiling(rpi[2]*N1),rmu[1],rsigma[1]), 
  rnorm(ceiling(rpi[3]*N1),rmu[2],rsigma[2]), 
  rgamma(ceiling(rpi[4]*N1),shape = 1,scale = rbeta[2]));
chr2 <- c(-rgamma(ceiling(rpi[1]*N2),shape = 1,scale = rbeta[1]), 
  rnorm(ceiling(rpi[2]*N2),rmu[1],rsigma[1]), 
  rnorm(ceiling(rpi[3]*N2),rmu[2],rsigma[2]), 
  rgamma(ceiling(rpi[4]*N2),shape = 1,scale = rbeta[2])); 
chr3 <- c(-rgamma(ceiling(rpi[1]*N2),shape = 1,scale = rbeta[1]), 
  rnorm(ceiling(rpi[2]*N2),rmu[1],rsigma[1]), 
  rnorm(ceiling(rpi[3]*N2),rmu[2],rsigma[2]), 
  rgamma(ceiling(rpi[4]*N2),shape = 1,scale = rbeta[2]));
# analyzing only chromosome 1 and chromosome 3
data <- list(chr1,chr3);

# fit GNG model with 2 normal components
test <- gng.fit(data, K = 2);
# vector of classification. 1 represents differential, 0 denotes non-differential
gngClass <- test$class;
# }