[InterMine Dev] /service/regions/sequence

Joe Carlson jwcarlson at lbl.gov
Fri Apr 10 02:33:05 BST 2015


While debugging some of our code I noticed that you're using the service 
URI /service/regions/sequence to trigger GenomicRegionFastaServlet 
instead of the servlet in earlier releases 
GenomicRegionSequenceExportServlet. The 'FastaServlet is also triggered 
by the url pattern /service/regions/fasta. I was curious why you've 
changed this. We're using the GenomicRegionSequenceExportServlet 
internally here. If this is going to be a permanent switch, then I'll 
probably want to make a new url pattern so that we don't have to worry 
about code merging issues in the future.

While we're on the subject, I'd actually like to extend the 
GenomicRegionSequenceExportServlet so that we can use it for dynamically 
generating gene upstream/downstream flanking sequence. This is easy 
enough to code up, but I wanted to discuss possible syntax for the 
region specification in case this can be incorporated into your code. 
gbrowse takes regions specifications like gene:<gene> and the like. I 
was going to GenomicRegionSearchUtil.parseRegion that could take a 
sequence feature to find the location. For our needs, we'd want to make 
it a little more pathquery-ish and be able to say 
Gene.secondaryIdentifier:<name> or something like that.

Next, do you have ideas on how you'd implement the extendedStart and 
extendedEnd in the regions specification syntax?

Just wondering if you had thoughts on this.



More information about the dev mailing list