Volcano plot from Posterior probabilities instead of p-value
Hi !
I'm currently diving into isoforms differential expression analysis results. I use the duo RSem-EBSeq to perform this work. I'd like to build a volcano plot of my data.
RSem-EBseq returns posterior probabilities (PP) and no p-values (PV) at all. I am aware, that both PP and PV allows me to answer the same question. I know that PP goes from 0 the lowest probability to 1, the highest. So I could just multiply those values b -1 to have a volcano plot in the same orientation as a "classic one".
Is it acceptable to do such a thing ?
• 3,042 views
•
link
written
by
Thibault D.
0 answers
No answers yet.
Log in to answer this question.
More posts like this
-
Tutorial: Unsupervised clustering on gene expression data
written by Hamid Ghaedi
-
heatmaps of RNA-seq data: what are accepted normalizations?
written by bwio •
-
Tutorial: Enrichment Analysis, Clustering and Scoring with pathfindR
written by egeulgen
-
EBSeq explained to a biologist
written by Rik Verdonck •
-
Understanding MMDiff output files
written by Thibault D.
It sounds valid to me to use the PP, but I'm not sure what you mean whit "classic one" and orientation.
If I do not multiply by -1, the volcano plot with PP would be reversed horizontally: the maximum value would be 0, the worst PP