R: A likelihood ratio test of population Hardy-Weinberg...
hwe.cc
R Documentation
A likelihood ratio test of population Hardy-Weinberg equilibrium for case-control studies
Description
A likelihood ratio test of population Hardy-Weinberg equilibrium for case-control studies
Usage
hwe.cc(model, case, ctrl, k0, initial1, initial2)
Arguments
model
model specification, dominant, recessive
case
a vector of genotype counts in cases
ctrl
a vector of genotype counts in controls
k0
prevalence of disease in the population
initial1
initial values for beta, gamma, and q
initial2
initial values for logit(p) and log(gamma)
Details
This is a collection of utility functions. The null hypothesis declares that the proportions of
genotypes are according to Hardy-Weinberg law, while under the alternative hypothesis, the expected
genotype counts are according to the probabilities that particular genotypes are obtained conditional
on the prevalence of disease in the population. In so doing, Hardy-Weinberg equilibrium is considered
using both case and control samples but pending on the disease model such that 2-parameter multiplicative
model is built on baseline genotype alpha, alpha beta and
alpha gamma.
Value
The returned value is a list with the following components.
Cox
statistics under a general model
t2par
under the null hypothesis
t3par
under the alternative hypothesis
lrt.stat
the log-likelihood ratio statistic
pval
the corresponding p value
References
Yu C, Zhang S, Zhou C, Sile S. A likelihood ratio test of population Hardy-Weinberg equilibrium
for case-control studies. Genetic Epidemiology 33:275-280, 2009
## Not run:
### Saba Sile, email of Jan 26, 2007, data always in order of GG AG AA, p=Pr(G),
### q=1-p=Pr(A)
case=c(155,27,4)
ctrl=c(408,55,15)
k0=.2
initial1=c(1.0,0.94,0.0904)
initial2=c(logit(1-0.0904),log(0.94))
hwe.cc("recessive",case,ctrl,k0, initial1, initial2)
### John Phillips III, TGFb1 data codon 10: TT CT CC, CC is abnormal and increasing
### TGFb1 activity
case=c(29,78,13)
ctrl=c(17,28,6)
k0 <- 1e-5
initial1 <- c(2.45,2.45,0.34)
initial2 <- c(logit(1-0.34),log(2.45))
hwe.cc("dominant",case,ctrl,k0,initial1,initial2)
## End(Not run)