Entering edit mode
Alasaari Jukka
▴
10
@alasaari-jukka-6176
Last seen 10.5 years ago
Hello,
I have Illumina 450k methylation data sets for 20 males (10 cases and
10 controls) and 20 females (10 cases and 10 controls), which I have
been analyzing separately using MINFI and LIMMA. The phenotype that
defines the cases and controls is the same for both data sets. Is
there any way to combine both data sets into one and analyze
differential methylation using sex as a covariate (with LIMMA)?
Here is the script that I am using:
dat <- read.450k.exp(base=basedir)
targets <- read.table("target.txt", sep="\t", header=T, as.is=T)
swan <- preprocessSWAN(dat2)
M <- getM(swan)
des <- model.matrix(~-1+targets2$Status)
colnames(des) <- c("CASE","CONTROL")
cont <- makeContrasts(CASE-CONTROL, levels=des)
fit <- lmFit(M, des)
fit2 <- eBayes(contrasts.fit(fit, cont))
vennDiagram(decideTests(fit2, p.value=0.001, adjust.method="BH"))
Thanks a lot in advance!
Best regards,
Jukka MSc
[[alternative HTML version deleted]]