I am using egdeR for differential analysis. It is between gene TULP4 silenced cell-line vs controls. As the sample size is very low 4 shTULP4 and 4 Controls, (shTULP4 vs Controls), I'm selecting differentially expressed genes only based on FDR < 0.05.
In this analysis I'm very much interested in one gene
CENPH. I expected this to be downregulated in shTULP4 samples. And I see this is downregulated, but FDR is not < 0.05. But p-value is < 0.05.
GeneSymbols logFC logCPM F PValue FDR CENPH -1.581838087 0.068915541 4.826724432 0.039121118 0.250871972
My question is:
Is it fine to say this gene is significantly downregulated in shTULP4 samples based on logFC and PValue? (OR) I have to go with only with logFC and FDR?
And for the volcano plot is it ok to take -log10(PValue) on y-axis instead of -log10(FDR) ?