Improve Research Reproducibility A Bio-protocol resource
- Home
- Protocols
-
Hello, I am trying to generate graphs similar to figures 3 and...
GA
G A Aug 10, 2021
Hello, I am trying to generate graphs similar to figures 3 and 4, but because I have a huge number of go terms (~300; (P<0.005, FDR <0.01)), I end up getting very crammed plots. In what ways could I change the script to suit my needs?
Sorry If my question sounds stupid; I do not unde Read more
Sorry If my question sounds stupid; I do not unde Read more
2
Views
Answer
Recommend
Follow
Share
1
answer
Sort by: most helpful / most recent
Titouan Bonnot Author Answered Aug 12, 2021
INRAE - Dijon, Agroecology
Hello,
I believe that you performed the GO term enrichment analysis powered by Panther, which gave you this high number of significant GO terms. You probably have a lot of redundancy between all these GO terms (GO terms associated with similar biological pathways). As described in our protocol, one way to reduce the redundancy between these GO terms is to use the tool REVIGO. In the protocol, we applied a stringent cutoff (dispensability < 0.05) to only keep and plot the highly representative GO terms. I would suggest something similar for your analysis. If you compare the enrichment between multiple groups of genes, please keep in mind that (as written in our protocol): "Several GO terms can be enriched in multiple groups but are not necessarily representative in all of them, based on REVIGO results and the applied cutoff."
Thanks,
Titouan
I believe that you performed the GO term enrichment analysis powered by Panther, which gave you this high number of significant GO terms. You probably have a lot of redundancy between all these GO terms (GO terms associated with similar biological pathways). As described in our protocol, one way to reduce the redundancy between these GO terms is to use the tool REVIGO. In the protocol, we applied a stringent cutoff (dispensability < 0.05) to only keep and plot the highly representative GO terms. I would suggest something similar for your analysis. If you compare the enrichment between multiple groups of genes, please keep in mind that (as written in our protocol): "Several GO terms can be enriched in multiple groups but are not necessarily representative in all of them, based on REVIGO results and the applied cutoff."
Thanks,
Titouan
0 helpful
0 unhelpful
0 comments ShareMy answer
Write your answer...
References (optional)
Add more
Post an Answer