Same Values In Limma/Microarray Differential Expression

Question

[Wan Fahmi]

Hello, I run limma for the differential expression of microarray microRNA data. I just wondering why am I getting the same value for the top list at least for 20 - 50 probeset. Is this a weird output from limma?

The code as below:

dat <- read.celfiles(list.celfiles("raw_data_dir"))
eset <- oligo::rma(dat)
design <- model.matrix(~0+Exp)
colnames(design)
fit <- lmFit(eset, design)
fit <- eBayes(fit, trend=TRUE, robust=TRUE)
results <- decideTests(fit, adjust.method="BH",p.value=0.05,lfc=2)

Here is the output of limma:

> topTable(fit, coef=NULL, number=10, genelist=fit$genes, adjust.method="BH",
+          sort.by="B", resort.by=NULL, p.value=1, lfc=0, confint=FALSE)
                ExpEarlyOnset ExpLateOnset  AveExpr        F      P.Value    adj.P.Val
MIMAT0002177_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0003130_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0006344_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0008160_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0009329_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0013186_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0013886_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0014943_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0015904_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51
MIMAT0023967_st      14.67058      14.6916 14.67923 122330.9 2.027378e-54 6.700116e-51

[Ram]

This should be a Question, not a Forum discussion. I've made the changes now, but please be more mindful in the future.

[Kevin Blighe]

There is a previous answer HERE; however, in your case, both the p- and adjusted p-values are the same.

I have seen this in the past with datasets of low sample n. There could also be an issue with the probe design on the microarray that you're using.

One question: why are you using trend=TRUE and robust=TRUE?