R package for power analysis for eQTL study
== TOADD ==
- Below is example code using the package to generate the similar power analysis curve as GTEx study1
# library(devtools);install_github("sterding/powerEQTL")
library('powerEQTL')
# sample size
N <- c(50,100,150,200,250,300)
nn <- length(N)
# MAF
MAF <- seq(0.5,20,0.1)/100
nq <- length(MAF)
# number of SNPs tested (10 SNPs per gene x 24 genes in total)
nSNP=240
# significant level (FP)
a=0.05
# obtain power
power_unbalanced <- array(numeric(nn*nq), dim=c(nn,nq))
for (i in 1:nn){
for (j in 1:nq){
# unbalanced
result <- powerEQTL.ANOVA(MAF=MAF[j],
typeI=a,
nTests=nSNP,
myntotal=N[i],
mystddev=0.13,
deltaVec = c(0.13, 0.13),
verbose = F)
power_unbalanced[i,j] <-result;
}
}
# set up graph
xrange <- range(MAF*100)
yrange <- c(0:1)
colors <- rainbow(length(N))
plot(xrange, yrange, log='x', type="n",
xlab="MAF (%)",
ylab="Power",
main="Power Estimation for eQTL Studies\nSig=0.05, nSNP=240 (one-way unbalanced ANOVA)")
abline(v=0, h=seq(0,1,.1), lty=2, col="grey89")
abline(h=0, v=c(1:10), lty=2,col="grey89")
# add power curves
for (i in 1:nn){
lines(MAF*100, power_unbalanced[i,], type="l", lwd=4, col=colors[i])
}
legend("topleft", title="Sample size (n)", as.character(N),fill=colors, bty='n')
Footnotes
-
Lonsdale J and Thomas J, et al. The Genotype-Tissue Expression (GTEx) project. Nature Genetics, 45:580-585, 2013. โฉ