lfcShrink MA plot comparison

Question

[user230613]

Hello, I have just started to use DESeq2 and I am trying to compare the results obtained with and without applying lfcShrink. These are the two MA plots, one with lfcShrink and one without:

Is it normal to lose all the significant (blue) genes with negative log2FC when applying shrinkage?

I can provide the commands if necessary but are the regular ones provided in DESeq2 vignette.

EDIT:

Summary of the results with shrinkage:

summary(resLFC)

out of 33014 with nonzero total read count
adjusted p-value < 0.1
LFC > 0 (up)       : 15, 0.045%
LFC < 0 (down)     : 33, 0.1%
outliers [1]       : 586, 1.8%
low counts [2]     : 10038, 30%
(mean count < 4)

Summary of the results without shrinkage: summary(res)

out of 33014 with nonzero total read count
adjusted p-value < 0.1
LFC > 0 (up)       : 11, 0.033%
LFC < 0 (down)     : 37, 0.11%
outliers [1]       : 586, 1.8%
low counts [2]     : 10038, 30%
(mean count < 4)

Snippet of the commands (nothing special, just regular commands of DESeq2 afaik):

sampleTable <- data.frame(condition = samples$Responder)
dds <- DESeqDataSetFromTximport(txi, sampleTable, ~condition)
keep <- rowSums(counts(dds)) >= 10
dds <- dds[keep,]
dds = DESeq(dds)
resLFC <- lfcShrink(dds, coef=resultsNames(dds)[2], type="apeglm")

[Michael Love]

You should always link to cross posts

https://support.bioconductor.org/p/9142791/