RNAseq GO enrichment analysis: GO organism .db

Question

[Neuls]

Hello,

I am analyzing RNAseq data for the first time. It consists of two bacterial strains of the same species cultured both in 5 different conditions. Briefly, I trimmed fastq sequences and I mapped the reads against a reference genome (Hisat2). Then, the transcripts were annotated and quantified (StringTie). Raw gene counts were extracted using a python script provided by the authors of these packages. Differential gene expression was assessed through DEseq2.

Now, I would like to perform a GO enrichment analysis within R environment. I have seen various packages, such as GOseq, topGO and GOstats. However, it seems they require a database containing the ontologies of each gene for a given organism (p.e. org.Mm.eg.db for mus musculus).

I downloaded the GO annotations for my organisms from QuickGO online resource, which contains the gene name/locus linked to GO ontologies. This is in tab-format plain text but there is also the option of retrieving this information in .gaf and .gpad format (I do not know their use).

My question is how to create a .db object like org.Mm.eg.db from ontology annotations contained in QuickGO database to perform GO enrichment analysis.

Thank you very much,

Answer 1

[e.rempel]

Hello,

my answer will be based on topGO R-package and approach used in this post

1. Using the information from QuickGO you could create a following list: names of the entries are the names of the genes, the entries are all the GO groups associated with the particular gene. In the post above this list is called geneID2GO.
2. Then you can follow my answer within the mentioned post: create the topGOdata object, run the tests etc.

HTH

[Neuls]

Thank you! Actually, I was digging more into topGO. The link to the post doest work, it would be great if you can fix it to run some tests as a proof of concept!

[GenoMax]

Neuls : I fixed the link in e.rempel 's answer. It will work now.