Functional annotation of B. subtilis

Question

[mk.applebee]

I have RNAseq data on B. subtilis, and mapped it to the genome so I have gene names and even BSU numbers for all of the genes. So far so good. Now I am struggling to find and assign functional annotations to each, preferably GO annotation.

Despite B. subtilis being a model organism, it is not included in the GO consortium database; they only have E. coli.

KEGG and bsubcyc both have B. subtilis functional annotations, but I cannot find a way to download any flat file from either site that associates each gene with its functional assignments.

How do people do this? What set of bioinformatics skill am I totally unaware of but that seem to be taken for granted? Do people "webpage scrape" for this info from KEGG? Is there a database format I need to learn?

Answer 1

[Lars Juhl Jensen]

The most straight-forward approach to obtaining GO annotation would in my opinion be to fetch it from UniProt. You can do that in several ways, for example, by querying for it via their REST API:

http://www.uniprot.org/uniprot/?query=taxonomy:224308&columns=id,go-id&format=tab

Answer 2

[Philipp Bayer]

Normally, I would use Blast2GO for this task, it compares your blast results of your species against known GO terms for these blast results and transfers them over. However, they changed their business model quite a lot - the command line version now costs at least 1600€ per year.

Then there's DAVID, which I have used to check for any enriched GO-terms, but looks like you can use it for "just" functional annotation too. In your case you might try the known B. subtilis IDs, or you could put the gene IDs of their blast hits from Swissprot/NR into DAVID and see what happens.

There's also STRAP which I haven't tried yet, here's the paper: Software Tool for Researching Annotations of Proteins (STRAP): Open-Source Protein Annotation Software with Data Visualization

[Lars Juhl Jensen]

Regarding using DAVID, beware that their backend data have not been updated since 2009 (https://david.ncifcrf.gov/content.jsp?file=update.html)

Answer 3

[Juke34]

You may also try interproscan

A: Get Associated GO terms to a transcript

Answer 4

[Stefan]

If you would like to generate functional annotations for your RNAseq data with Blast2GO you can (1) blastx your sequences against e.g. E.coli or NR and then (2) proceed with the Gene Ontology mapping and (3) function assignment/annotation. Just make sure that you have proper transcripts before starting the blastx.

Once you have GO terms for your sequence you can also link to KEGG via enzyme codes and use InterProScan and RFAM to further refine your results.

You can use the free Blast2GO Basic version for this and no special bioinformatics skills are required.

You may also have a look at the Blast2GO BioMart Import function to import functional annotations if you already have gene names or IDs. Not sure about B.subtilis and BioMart so.