Visualizing Annotation Results Using R?
1
0
Entering edit mode
5.9 years ago

I have a gene list that I want to deem "special" or "unspecial". What typically is done (to my understanding) for this is a  gene annotation analysis (like DAVID). My question is how are these analyses:

1. automated, so results are visualized dynamically, instead of via manual entry for each gene list on the DAVID website, etc.

2. visualized in a compelling way (are there just tables with the annotation terms that are significant?)

3. done in R

So far, I have discovered clusterProfiler, but its enrichDAVID() method seems to be deprecated, and I only have official gene symbols, not ENTREZ gene ids. I'm hoping someone has a working (and maintained) way to do these analyses in R.

Thanks for any help!

annotation R visualization gene • 2.0k views
2
Entering edit mode
5.9 years ago

Hi,

If you don't like the enrichDavid function, you can use the enrichGO and groupGO from clusterProfiler. They work on the same database (GeneOntology) and they actually provide more updated results. The usage is the same, just specify change the name of the function.

To convert gene IDs to symbols, you can use the Bioconductor way of doing so:

> library(AnnotationDbi)
> library("org.Hs.eg.db")
# Converting symbols to entrez. Watch out for NAs and One-to-Many mappings!
> AnnotationDbi::select(org.Hs.eg.db, keys=c("MGAT3", "MOGS", "PTEN"), columns="ENTREZID", keytype="SYMBOL")


In my experience it is always better to convert all your data to either Entrez or Ensembl Ids before starting any analysis. Working with symbols is tricky because the same symbol can map to multiple ids and this can lead to headaches later on.