duplicate isoforms in chain_samples.py output？ about cdna_cupcake

[Magdoll]

Hi @52teth ,

This is a known "issue" with chaining single-exon transcripts. While chaining -- establishing "equivalence" of multi-exon isoforms is much more straightforward (currently, I simply ask that all internal junctions agree), for single-exon transcripts, I currently just ask that they overlap.

This creates the weird case in which your PB.1.1 from sample 1 is a single exon of chr1:70-4943 but you have two separate transcripts in sample 2 PB.1.1 and PB.2.1 that are non-overlapping but are actually two single-exon transcripts that are each a fragment of sample 1's PB.1.1!

So, by the current logic of chaining, the result is correct. Because PB.1.1 in sample1 is by current definition equivalent to both PB.1.2 and PB.2.1 in sample 2.

Hope this makes sense.

--liz