Differentiating between Constitutively spliced and Alternatively spliced exons

Question

[dally]

Hey guys,

So this is kind of just a curious question. I was wondering what software and methodology is out there and if it's possible to take a reference genome of say hg38 for instance (from UCSC ?) And first ... give me a database or file of all the alternatively spliced exons (isoforms) ... and then differentiate whether an exon isoform is constitutive or alternative.

I'm aware the question is rather vague, but it's really just a thought and I don't have any real questions behind this. Just wondering if it's more of a "gotta create your own custom script" or if there's some tools out there to do this and is thereby possible for someone with little programming knowledge.

Answer 1

[Devon Ryan]

Depending on how exact you want to be with your definition of an exon you might be able to do this with cut and uniq. I suspect, though, that this would be reasonably doable in R with GenomicRanges (load the GTF file, split by gene, pass that into a function that (1) makes disjoint intervals, (2) counts the intersects of them with the original GRanges object and (3) returns those ranges with more than 1 count).

Obviously the results of this are limited by the quality and completeness of the annotation, so don't read excessively into the results.