Entering edit mode
Greetings,
I have been doing some RNA analysis with the limma package using the
eBayes
function. I have a matrix of counts consisting of 25322 genes and 23
samples. I intended to use the voom function to normalize my data.
Here
is my code and problem:
design <- model.matrix(~factor(class))
colnames(test) <- c("Fasted", "Prefed")
nf = calcNormFactors(sub, method = "TMM")
voom.data <- voom(sub, design = design,
lib.size = colSums(sub) * nf)
voom.data$genes = rownames(sub)
voom.fitlimma = lmFit(voom.data, design = design)
voom.fitbayes = eBayes(voom.fitlimma)
Warning message:In ebayes(fit = fit, proportion = proportion,
stdev.coef.lim = stdev.coef.lim, :
Estimation of var.prior failed - set to default value
I tried looking into the error and couldn't find a clear answer in the
archives (but I may have missed it). What I gather is one of the
estimated
parameters couldn't be and a default was entered instead. I don't
think I
should just fall on a default when there seems to be something that
should
be solvable. Curious I decided to do subsets of my data and the
analysis
would run up to 23800 rows but not 23801 and up. I then tried to use
rows
10000:25322 and that works perfectly!?!?
Is there some sort of limit on rows for eBayes? - I highly doubt it
but I
need to ask given what I am running into.
Any thoughts would be sincerely appreciated as I am vexed by this.
Thanks for your time,
Sincerely,
--
Charles Determan
Integrated Biosciences PhD Student
University of Minnesota
[[alternative HTML version deleted]]