How to perform functional gene clustering?

Question

[ThePresident]

Let's say I have a gene "A" and I want to know what is a genomic context around it in different bacterial genomes. So, I have looked around "A" in bacterial genome X, Y, Z etc. and I have pulled out a few genes that are in vicinity (both upstream and downstream) of "A" for each one of these X-Y-Z genomes.

Now, I have a list containing all these genes with their locus_tag and functional description. Based on these (and I have a lot of them), I would like to see if any specific functions are enriched around gene "A", i.e. look for clusters of orthologs (COGs) in these genes surrounding "A".

How this can be done? Initially, I wanted to use DAVID but it seems that it's main function is to look for enrichment in differentially expressed genes following RNA-seq or similar experiment. Also, my list of genes is coming from a number of various organisms.

Thank you, TP

[Buffo]

What about ortoMCL, or ortoMCL-DB?

[ThePresident]

Thanks for the suggestions, but I'm afraid these won't do (at least based on the description they give you and the fact that I have list of locus_tags from >200 different species).

Answer 1

[Kevin Blighe]

ThePresident, I believe that DAVID is for any type of enrichment on any group of genes. It is a reasonable place to start with your idea for exploring COGs (which sounds like a very interesting project, by the way). I would be interested in GO terms and proteins domains, mostly.

There is another thread here on gene enrichment (in R): GO enrichment analysis using R

If you are looking at GO terms, you should also take a look at my answer here about how to interpret these: A: Go annotation reliability ?

[ThePresident]

Thank you for clarification. I'll give DAVID a try and see what I get because, as you mentioned, it is an enrichment after all. What I like about DAVID is that it takes locus_tags and thus can deal with anything that has been annotated on NCBI. Also, there seems to be a nice R package for DAVID that I'd like to use.

Thanks again :)

[Kevin Blighe]

Yes, DAVID accepts many types of gene annotation, even prove IDs from microarrays.

Best of luck!