Question

Limma Ebayes on SILAC data

0

Entering edit mode

reventropy2003 ▴ 10

@reventropy2003-14321

Last seen 6.2 years ago

Is there any need to set up a design or contrast matrix using limma ebayes on ratiometric replicates? I have three replicate ratios that are log2 transformed. I'm running ebayes:

data = read.csv(paste(inpath,fname,sep=""),header=T,stringsAsFactors=F,na.strings="NaN",sep="\t")

fit = lmFit(data[,c("rep1","rep2","rep3")]))
fit2 = eBayes(fit)
tt = topTable(fit2, adjust="BH",number=Inf,confint=T,sort.by='none')

Where "rep 1-3" are the columns containing the ratio measurements. Is this correct?

Thanks,

Jeremy

ebayes limma bioconductor • 1.7k views

ADD COMMENT • link updated 7.2 years ago by Gordon Smyth 52k • written 7.2 years ago by reventropy2003 ▴ 10

0

Entering edit mode

Hi Jeremy,

I meet the same problem, I also want to use limma package to deal with my SILAC data, but I really confused what codes should I use. Should I build a design matrix?

If you could reply me, so appreciated it!

Thanks so much! Best, Peiru

ADD REPLY • link 6.0 years ago chenpeiru1994 • 0

score 1 · Answer 1 · 2017-11-04

I am not familiar with SILAC ratiometric data but, if it produces log-expression ratios, then you could analyse the results like those from two colour microarrays. Your data appears to be similar to the design considered in Section 11.2.1 of the limma User's Guide. In that case, you don't need to create a design matrix -- the default linear model with just an intercept will do the job -- so your code seems correct.