Integrating gene ontology DB's into pipeline scripts

Question

[joemrosen]

Does anybody have suggestions for integrating gene ontology DBs or BLAST/refseq into bioinformatics pipelines? My goal was to be able to automatically search my output (filtered by impact refseq transcript ID's and/or HUGO gene IDs) against a gene ontology DB such as DAVID or Refseq, then automatically fetch the results to display in a clinical report. I was thinking of maybe building a web scraping script since there doesn't seem to be much support for this sort of workflow API wise.

[Pierre Lindenbaum]

why scraping as all those databases are available for download ?

Answer 1

[EagleEye]

Have a look at GeneSCF. It is easy to integrate in any bioinformatics pipelines.

[joemrosen]

Looks good, thanks for the tip!

Answer 2

[Istvan Albert]

There should be no need to scrape websites; there are a large number of tools that implement various ontology and functional annotation-related functionalities.

Most are written in R. I would investigate those.

FWIW even DAVID has a web API:

• https://davidbioinformatics.nih.gov/content.jsp?file=WS.html
• https://davidbioinformatics.nih.gov/content.jsp?file=DAVID_API.html

[joemrosen]

I tried the DAVID api but the output isnt very detailed, I'm not looking for simple gene annotation, I already have that with snpEff and other tools, I was looking to write a script that takes my list of annotated genes, queries the DB for deep ontological info, and generates an in depth report based on it. some thing like the example below, I know blast has rBlast, but it does it provide deep detail?

High Impact Variant Results: NM_123: High impact, associated with X disease etc.. NM_456: High Impact associated with Y disease, seen in Z populations etc