Error in DEqMS proteomics analysis

Question

[Mrinal.singh06]

Hi, I have the raw proteomics data and I am doing differential analysis using DEqMS.

I am using two variables to create a group

groups<factor(paste(final_val_C$Qp_Group,final_val_C$Day,sep='.'))

The day is numeric and Qp_Group consists of "high" and "low variables. Days consist of 3,5,8 and 11 days

My group combination looks like this

high.3 high.5 high.8 high.11 high.3 high.5 high.8 high.11 high.3 high.5 high.8 high.11 high.3 high.5 high.8 high.11 high.3 high.5 high.8 high.11 high.3 high.5 high.8 high.11 low.3 low.5 low.8 low.11 low.3 low.5
low.8 low.11 low.3 low.5 low.8 low.11 low.3 low.5 low.8 low.11 low.3 low.5 low.8 low.11 low.3 low.5 low.8 low.3 low.5 low.8
low.11 low.3 low.5 low.8 low.11 low.3 low.5 low.8 low.11
Levels: high.11 high.3 high.5 high.8 low.11 low.3 low.5 low.8

design = model.matrix(~0+groups)
colnames(design) = gsub("groups","",colnames(design))
fit1 <- lmFit(log2(proteomeRaw), design)
fit3 <- eBayes(fit1)
EqMS.results = outputResult(fit3)

I am not using contrasts as I don't have any separate ones to compare. My outputResult is throwing this error even when I use Day as char.

"Error in `$<-.data.frame`(`*tmp*`, "sca.adj.pval", value = numeric(0)) : replacement has 0 rows, data has 8183"

Any idea

Answer 1

[Gordon Smyth]

You have to use contrasts. You have 8 different groups to compare. The software can't guess for you what comparisons you want to make!

You are actually not using DEqMS at all. DEqMS is a wrapper on limma anyway, but since all the analysis here is straight limma rather than DEqMS, you must use straight limma output functions like topTable(). The DEqMS wrapper function outputResult is out of place because you haven't done any DEqMS analysis.