I have the RNA seq data for the differentially upregulated and downregulated genes. The data has p values and FDR values for differentially regulated genes. Here, is the question do we need to apply FDR for DGE analyses and trim false-positive p values, even when we want to just show the difference between some selected gene of our interest?
Dear Lukasz,
yes, you should focus on FDR and not p values. The p values are not adjusted for multiple testing, you will get many "significant" hits that you will not find when looking at the FDR.
Hope this helps & with best regards,
Alexander
Hello Lukasz Kuryk,
I agree and agree with Alexander Neumann on the use of unmodified p values. However, it may be interesting to combine several evaluation factors (e.g., negative log10 of adjusted p values, log2 Fold Change, etc...) for the evaluation of differentially expressed genes (DEGs).
However, the Benjamini and Hochberg's False Discovery Rate (FDR) correction -particularly the modified version described by Storey (that produces q-values)- has become invaluable in transcriptional profiling, and large-scale bioinformatics analysis in general. At its simplest level, the FDR takes a list of p values, and sorts them in order from smallest to largest.
However, p-values (and FDRs) have their own problems. At small p values, they can be twitchy from study to study, and show big swings while fold changes, particularly for well-annotated genes with reliable signal intensity/counts, can be much more stable across studies.
So generally, an amalgam of fold change and p value, especially the Volcano plot, can be used to identify reliable, strong expression changes.
Hoping that this comment will be useful. Feel free to discuss or share some informations about your RNA Seq data studies.
Ludovic Dumont, Ph.D.
- Badges
- Science topic
- Similar topics
- Medicine
- Disease
- Infectious Diseases