Gene Calling

[charlie]

Hey everybody,

I have a rather stupid question. How do people commonly associate known genes with large numbers of mapped regions in the genome? I.e. I have large numbers of chromosomal locations (something like "chr42 1313131313 +") and want to know which (if any) known gene this corresponds to. This is such an obvious problem that everybody here must be facing, yet I haven't been able to find a tool or anything to quickly do that. Of course, I could just write a script to do that, I suppose, but I'd like to save myself some work :-).

Thanks!

-Charlie

[apfejes]

Hi Charlie,

I have a very simple class called FindFeatures that I'm working on, which is available as open source code (http://vancouvershortr.sourceforge.net). Because it's relatively early in the development cycle, there isn't much documentation for it yet (though, if there was interest, I could change that.)

Essentially, the code uses the Ensembl database to find genes which overlap with regions of interest - no matter how big or small that overlap is, at the moment.

If you know java, you could quickly figure out the command lines from the code. Otherwise, I could put together some instructions, if you're interested.

Anthony

[chris]

For higher organism you can use the perl API in Ensembl, it's very powerful and easy to use. You can simple give it a chromosomal region like you have and it will tell you which genes it overlaps with. I like it lots.

http://www.ensembl.org/info/data/api.html

For lower organisms I don't know of an equivalent resource; I'd love to know if there is one!

[apfejes]

Since Chris mentions it, I may as well be clear; the utility I proposed above is just a Java Ensembl API implementation, which is very similar to the perl one. (They both use the same database.) I'm not a big fan of perl for code re-use issues, but you'll get the same result.

[frozenlyse]

Or you can upload your dataset to UCSC genome browser (upload custom track) and then use the table browser to intersect your ROIs with various annotations

this should give you an idea of how to start