Hello everyone,
I am trying to use R to draw a volcanic plot after DEG analysis. In my case, I use samr for detect DEG. I have successfully used ggplot2 to draw it. However, I have trouble when I want to add gene names into my plot.
1st: I dont know how to convert all of the probe name into gene name. There are a way on google. But the problem is I can not know how to solve when 1 probe can be 2 genes or more.
2nd: I dont know what is Padj as a cut-off to add gene names into the volcanic plot. What is Padj? And how much for the cut-off?
Thank you so much!
Hi Phụng Dương
2nd question first. Padj is the adjusted p-values. Since you are running many-many tests at the same time, you should expect some p-values to be very small by chance. For that reason, the R package provide padj to account for multiple testing (read more here, https://en.wikipedia.org/wiki/Multiple_comparisons_problem). As of the cut-off, it depends on which multiple testing adjustment is used, but in general this cut off would be arbitrary. For example, when using FDR (false discovery rate) the cut-off could be 0.2 (very liberal, expect many false positives) to 0.05 (less so).
1st question. There are multiple ways to deal with genes represented by more than one probe. For example, you could average the probes, choose the highest variance or the highest expression level. My fav method in this regard is to use collapseRaws from the WGCNA package.
Adding names to a volcano plot, as in any other ggplot2 graph can be done using either 'geom_text()' or 'annotate()'. The first can map the names column to each point in the graph, the second can be used to add selected names manually to the graph.
Thank you so much for your answer. Now I understand the way to do more. I will try to figure out how to do it as you say.
hello,
I believe you understood padj value definition.
you can write the table of differential expressed genes by filtering them on padj
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R