generalized_poisson_likelihood: Maximum Likelihood Estimates for the Generalized Poisson...

In GPseq: gpseq: Using the generalized Poisson distribution to model sequence read counts from high throughput sequencing experiments

Description

This function calculates the Maximum Likelihood estimates for theta and lambda when vector y is fit to the Generalized Poisson Model. Newton Raphson Method is employed to calculate the MLE. The values are only valid if the Newton Raphson Method converges.

Usage

generalized_poisson_likelihood(y)

Arguments

y	Vector of counts.

Value

mark	1 if the Newton Raphson Method converges. If mark = 0, then the values of theta and lambda are not applicable
theta	Maximum Likelihood Estimate for theta in the Generalized Poisson Model(theta,lambda)
lambda	Maximum Likelihood Estimate for lambda in the Generalized Poisson Model(theta,lambda)
y_bar	Mean of y which is also the Maximum Likelihood Estimate for lambda for the Poisson model
length	Length of y which will be later used to calculate the normalization values

Author(s)

Sudeep Srivastava, Liang Chen

References

Consul, P. C. (1989) Generalized Poisson Distributions: Properties and Applications. New York: Marcel Dekker.
Sudeep Srivastava, Liang Chen A two-parameter generalized Poisson model to improve the analysis of RNA-Seq data Nucleic Acids Research Advance Access published July 29,2010 doi : 10.1093/nar/gkq670

Examples

y = rpois(100,10);
out = generalized_poisson_likelihood(y);
#Check if it converged
if(out$mark==1)
{
#Value of Theta
  cat("theta = ",out$theta,"lambda = ",out$lambda,"lambda_poisson = ",out$y_bar,"Length = ",out$length,"\n");
}

GPseq documentation built on May 30, 2017, 3:11 a.m.