Shaun Jackman

Here's a related issue: https://github.com/genometools/genometools/issues/225 and a useful reference: http://sequenceontology.org/so_wiki/index.php/Discontinuous_features

@gordon wrote… https://github.com/genometools/genometools/issues/225#issuecomment-34348838 > I also don't understand how one would translate such a CDS into a protein sequence. Does anyone understand what the spec says about this? My guess...

For future folk who may stumble on this issue, here's my workaround for now to extract and translate trans-spliced genes that span multiple sequences. I name the sequences `Name=gene_part1` `Name=gene_part2`...

I'm interested in trans-spliced genes as well.

:+1:

@yeban Bump?

No worries, Anurag. I don't have time right now either. Perhaps in about a month.

It's a weird edge case, but yes it's the standard way of representing a trans-spliced gene. See http://sequenceontology.org/so_wiki/index.php/Discontinuous_features It's a very minor aesthetic thing (so very low priority), but I...

The CDS is also a multifeature. I was missing the `ID=CDS1` tag in my above example. I've added it.

Thanks! That worked like a charm. I think that would be a useful default behaviour, or perhaps the error message could suggest `-matchdescstart`. From the `--help` output, which I did...