Dear List,
I have been trying to use p.adjust() to do BH multiple test correction and have
gotten some unexpected results. I thought that the equation for this was:
pBH = p*n/i
where p is the original p value, n is the number of tests and i is the rank of
the p value. However when I try and recreate the corrected p from my most
significant value it does not match up to the one computed by the method
p.adjust:
> setwd("C:/work/Methylation/IMA/GM/siteLists")
>
> hypTable <- read.delim("hypernormal vs others.txt")
> pList <- hypTable$p
> names(pList) <- hypTable$site
>
> adjusted <- p.adjust(pList, method="BH")
> adjusted[1]
cg27433479
0.05030589
>
> pList[1]*nrow(hypTable)/1
cg27433479
0.09269194
I tried to recreate this is a small example of a vector of 5 p values but
everything worked as expected there. I was wondering if there is some subtle
difference about how p.adjust operates? Is there something more complicated
about how to calculate 'n' or 'i' - perhaps due to identical p values being
assigned the same rank or something? Does anyone have an idea what might be
going on here?
Many thanks,
Scott
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
