Hello teachers!
I have a quick question about DEG and p-value interpretation in RNA-seq data.
When I use edgeR to get DEGs, two values are reported: log2FC and p-value. As you may know, many researchers consider DEGs satisfying with |log2FC| > 0.58 and p-value < 0.01. Here is my questions:
1. p-value < 0.01 means "significance of difference" and |log2FC| means strength.. (am I right?).
2. When I see my data, there are half of genes showing |log2FC| < 0.58 around p-value 0.01 ! Do I interpret that p-value 0.01 is too large to select DEGs?
Thanks for your attention!
•
written
by
progistar •
The raw p-value is useless, you must look at a p-value which has been adjusted for multiple testing. You should also look up what p-values actually mean. You should also know that log2LFC means the log base 2 of the fold change.
There are no objectively correct cutoffs for those values.
Thanks for your reply. From your advice, I used FDR (adjusted p-value) to see the aspect; however, there are half of genes showing |log2FC| < 0.58 around FDR 0.01 still. Is it reasonable to Adjust FDR threshold until genes with |log2FC| > 0.58 are popular ?
Log in to answer this question.
