Hey, I want to perform a GO term enrichment analysis by comparing a list of affected protein to all protein identified and/or genome.
First I performed GO Slim mapper at SGD and then performed Fisher exact test on each GO term individually. After doing all of that I realized I should probably have compared affected proteins to all proteins identified and not the the entire genome.
I have noticed that a lot of papers use DAVID so I uploaded a gene list and the background list but my analysis kinda stops up. I get that there are different levels for GO Annotation, but I just do not know how to use them.
I want to know - does the affected proteins have a specific subcellular localization? - does the affected proteins share biological function? - does the affected protein interact with specific protein complexes?
Any help is highly appreciated!
Also, most papers presents their data with a bar graph that looks like it has been generated by the same source. Does anyone know how to do this?
Not an answer to your question but: don't use David. Outdated for years.
Instead, go for EnrichR or Webgestalt.
http://amp.pharm.mssm.edu/Enrichr/
http://www.webgestalt.org
The Webgestalt told me it moved to http://www.webgestalt.org/option.php :)
Thanks for notifying me, I updated my post.
I appreciate any feedback you may have ;)
Is there something in specific with DAVID that is out of date? Does EnrichR or Webgestalt allow me to compare two datasets (affects vs identified) or do they compare with genome frequency?
I am wondering if I am going to go with GO slim analysis through SGD, but "braiding" the data and doing the Fisher test will ages ;)
The datasets (genesets/annotation) from David are out of date. But somewhere I read an update was announced. There are multiple posts discussing gene enrichment analysis on Biostars, also comparing alternatives.
With regard to your question: enrichR does not allow you to specify a backgroundlist (and assumes thus the entire genome is measurable), webgestalt does allow you to give a backgroundlist of what you could identify.
WebGestalt it is ;) Hopefully it is more intuitive than DAVID.
Thank you!
Glad to help. I just saw there is also something in Webgestalt called Go View, maybe you want to give that a try. I have no idea what it does but it definitely sounds fancy. Will test it later this week. http://www.webgestalt.org/GOView/ After some more looking around, WebGestalt's database is apparently from 2013.
I would like to do something similar to Fig 2C in Henriksen et al 2012 MCP