Dear All,

I am working with the RNA-Seq of mouse data, I have a wild-type group(sample size=5), and another group that was knocked out for a gene (sample size=5).

here is the code I used:

dds <- DESeqDataSetFromHTSeqCount(sampleTable=sampleTable,
                                  directory=folder,
                                  design=~Sex+condition)

dds <- estimateSizeFactors(dds)
keep <- rowSums( counts(dds) >= 10 ) >= 5
dds <- dds[keep,]

colData(dds)$condition <- relevel(colData(dds)$condition, ref = "WT")

dds<- DESeq(dds) 

resultsNames(dds)

res<- results(dds, name = "condition_KO_vs_WT")

But as there is a clear biological difference between my groups; I have a lot of genes with extreme low padj values. ex of some of them (they are more):

baseMean    log2FoldChange  lfcSE   stat    pvalue  padj
2268.61812  -6.394621706    0.125791377 -50.83513568    0   0
1456.067732 -8.250365901    0.231033179 -35.71074045    2.69E-279   2.21E-275
1385.788289 2.067801423 0.077752666 26.59460487 7.84E-156   4.28E-152
482.3232806 -2.240098286    0.10594365  -21.14424308    3.12E-99    1.28E-95
9679.679616 1.135952663 0.078263158 14.51452633 9.80E-48    3.21E-44
1271.405292 1.021987101 0.071287705 14.33609209 1.30E-46    3.55E-43
79.04433843 4.369034503 0.319439806 13.67717619 1.39E-42    3.25E-39
741.2276181 1.247591601 0.10804284  11.54719366 7.63E-31    1.56E-27

which approach do I have to use to analyze my data Also, I used the greaterAbs which also I am not sure when I have to use it. (but by the way, it didn't change the results)

res <- results(dds, contrast=c("condition","KO","WT"), altHypothesis="greaterAbs",lfcThreshold = 1)

May you please guide me with the way how to analyze my data?

Thanks a lot!

If you follow the vignette you're good to go. I don't think there is more to say here other that you should do PCA to diagnose potential (nested) batch effects and that you can use lfcShrink with the lfc argument to test against a minimum fold change in case you have many DEGs and want to prioritize these with large reliable effect sizes.