genomic control based on p values
The function obtains 1-df chi-squared statistics (observed) according to a vector of p values, and the inflation factor (lambda) according to medians of the observed and expected statistics. The latter is based on the empirical distribution function (EDF) of 1-df chi-squared statstics.
It would be appropriate for genetic association analysis as of 1-df Armitage trend test for case-control data; for 1-df additive model with continuous outcome one has to consider the compatibility with p values based on z-/t- statistics.
gcontrol2(p,col=palette()[4],lcol=palette()[2],...)
p |
a vector of observed p values |
col |
colour for points in the Q-Q plot |
lcol |
colour for the diagonal line in the Q-Q plot |
... |
other options for plot |
A list containing:
x |
the expected chi-squared statistics |
y |
the observed chi-squared statistics |
lambda |
the inflation factor |
Devlin B, Roeder K (1999) Genomic control for association studies. Biometrics 55:997-1004
Jing Hua Zhao
## Not run: x2 <- rchisq(100,1,.1) p <- pchisq(x2,1,lower.tail=FALSE) r <- gcontrol2(p) print(r$lambda) ## End(Not run)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.