I've downloaded the CEL files from GSE154619 and want to run a DE test for multiple contrasts. In essence, this data has two treatments(drug vs control) in two tissues. Here's my code:

suppressPackageStartupMessages({
library(affy)
library(oligo)
library(limma)
library(pd.clariom.s.mouse)
library(clariomsmousetranscriptcluster.db)
library(biomaRt)
})

data_dir <- './data/DpQ/GSE154619/'

metadata <- data.frame(
  treatment = factor(ifelse(grepl('DQ', all_files), 'DQ', 'CTRL')),
  tissue = factor(ifelse(grepl('A', all_files), 'AF', 'NP'))
)
condition <- factor(paste0(metadata$treatment, '_', metadata$tissue))
rownames(metadata) <- all_files

pdata <- as(metadata, 'AnnotatedDataFrame')
affy_raw <- file.path(data_dir, list.files(data_dir)) |>
  read.celfiles(phenoData = pdata, sampleNames = sampleNames(pdata))

boxplot(affy_raw)

eset <- rma(affy_raw)

boxplot(eset)

design <- model.matrix(~ 0 + condition)
fit <- lmFit(eset, design)
contrasts <- makeContrasts(
  conditionDQ_AF - conditionCTRL_AF,
  conditionDQ_NP - conditionCTRL_NP,
  levels = design
)
fit <- contrasts.fit(fit, contrasts)
fit <- eBayes(fit)

summary(decideTests(fit))

The last line produces:

      conditionDQ_AF - conditionCTRL_AF conditionDQ_NP - conditionCTRL_NP
Down                                   0                                 0
NotSig                             29129                             29129
Up                                     0                                 0

Which isn't what I expect, there should be many significantly DE genes. The box plots look like:

Thanks in advance!