power.pga: Power and sample size for case-control genetics studies

In statapps/power.ctepd: Power and sample size for clinical trials and epidemiology

Power and sample size for case-control genetics studies

Description

Compute power and sample size of test or determine the relative risks (RR) under alternative hypothesis to obtain target power (same as power.anova.test).

Usage

  power.pga.snp(n1=NULL, cc.ratio=1, p0=0.01, maf=0.05, daf=0.05, Dp=NULL, 
              R2=NULL, rr=NULL, rr2=NULL, sig.level=0.05, power = NULL, 
              edf=1, genModel=c('co.dom', 'co.dom2', 'dom', 'rec'),
              tol = .Machine$double.eps^0.25)

  power.pga.hap(n1=NULL, cc.ratio=1, p0=0.01, dhf=0.05, n.hap=2, 
              rr=NULL, rr2=NULL, sig.level=0.05, power = NULL, 
              genModel=c('co.dom', 'co.dom2', 'dom', 'rec'),
              tol = .Machine$double.eps^0.25)

Arguments

n1	Number of cases
cc.ratio	Control / case ratio, default is 1
sig.level	Significance level (Type I error probability)
power	Power of test (1 minus Type II error probability)
Dp	D', default is 1, only specify one of Dp and R2
R2	R-square, default is 1
p0	Disease prevalence, default is 0.01
maf	Minor allele frequency, default is 0.05
daf	Disease allele frequency, default is 0.05
dhf	Disease haplotype frequency, default is 0.05
n.hap	Number of haplotypes, default is 2
rr	Relative risk (RR)
rr2	Second relative risk for co-dominant model
genModel	'co.dom': co-dominant mdoel (1 df), 'co.dom2', co-dominant model (2df), 'dom': dominant model, 'rec': recessive model
edf	edf: Effective number of tests, default is 1
tol	tol for uniroot function

Details

Calculate power and sample size of case-control study for genetics analysis (either SNP or haplotype) using a co-dominant, dominant or recessive disease susceptibility model. The algorithm is based on methods proposed by Menash, Ronenberg and Chen (2008).

Exactly one of the parameters 'n1','power' and 'rr' must be passed as NULL, and that parameter is determined from the others. Notice that the last one has non-NULL default so NULL must be explicitly passed if you want to compute it.

Value

Object of class 'power.htest', a list of the arguments (including the computed one) augmented with 'method' and 'note' elements.

case.control	A vector for number of cases and controls
D_max	Maximum disequilibrium D_max = min(p1(1-pd), pd(1-p1)), output for power.pga.snp
D	Non-scaled disequilibrium with D = Dp*D.max, or D = r*D.max, where D.max is the maximum disequilibrium, with D.max = min(p1(1-pd), pd(1-p1)). This is an output for power.pga.snp
n.total	Total number of sample size for the study
power	Power of the study
rr	Releative risk with one copy of disease allele or disease haplotype
rr2	Relative risk with two copies of diseaes allele or diseas haplotype
sig.adjust	Ajusted significant level based on number of tests for SNP analysis
sig.level	Significant level
Penetrances	The penetrance rates for 0, 1 and 2 copies of disease SNP or haplotype

Note

'uniroot' is used to solve power equation for unknowns, so you may see errors from it, notably about inability to bracket the root when invalid arguments are given.

Author(s)

Bingshu E. Chen ([e-mail])

References

Menashe, I., Rosenberg, P. S. and Chen, B. E., (2008). PGA: power calculator for case-control genetic association analysis. BioMed Central (BMC) Genetics. vol. 9, page 36.

See Also

power.t.test, power.surv.test, pwr.2p.test

Examples

# Example from paper by Menash, Ronenberg and Chen (2008).
power.pga.snp(power = 0.9, rr = 2.0, genModel = 'co.dom', p0 = 0.07, edf=200)
power.pga.hap(power = 0.9, rr = 1.7, genModel = 'co.dom', p0 = 0.07, n.hap=12)

statapps/power.ctepd documentation built on April 26, 2024, 3:22 a.m.