Hello, I have an issue with the differential enrichment analysis after data imputation (Introduction to DEP).

I have installed and loaded the limma package. However, I am unable to perform the differential enrichment analysis based on a linear model and empirical Bayes statistics (see photo). Does anyone have a solution here? Or is there something wrong with my data in this case?

> library(limma)
> # Test every sample versus control
> data_diff <- test_diff(data_imp, type = "control", control = "Lum1")
Tested contrasts: Lum2_vs_Lum1, Lum1_1_vs_Lum1, Lum2_1_vs_Lum1
Fehler in .ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim,  : 
  No finite residual standard deviations
> # Test all possible comparisons of samples
> data_diff_all_contrasts <- test_diff(data_imp, type = "all")
Tested contrasts: Lum1_vs_Lum2, Lum1_vs_Lum1_1, Lum1_vs_Lum2_1, Lum2_vs_Lum1_1, Lum2_vs_Lum2_1, Lum1_1_vs_Lum2_1
Fehler in .ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim,  : 
  No finite residual standard deviations
> # Test manually defined comparisons
> data_diff_manual <- test_diff(data_imp, type = "manual", 
+                               test = c("Lum1", "Lum2"))
Tested contrasts: Lum1, Lum2
Fehler in .ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim,  : 
  No finite residual standard deviations

The error message from limma says that the data consists of infinite values, so there seems to be something fundamentally wrong with your data_imp object.

I note however that you are running the DEP function test_diff rather than running limma directly, and I don't know anything about the DEP package.

What do you get if you type

summary(assay(data_imp))

?