[Biopython-dev] LocationParserError

[Peter Cock]

On Fri, Mar 9, 2012 at 7:53 AM, Matthias Bernt <MatatTHC at gmx.de> wrote:
> Hi,
>
> I just got the new RefSeq 52 release and found a really strange error
> causing an exception:
>
> Bio.GenBank.LocationParserError: join(complement(149815..150200),complement(293787..295573),NC_016402.1:6618..6676,181647..181905)
>
> the accession is: NC_016406. Any ideas?
>
> Matthias

That is the most complicated trans_splicing feature I've seen in a while :)

It says gene nad1 is (trans) spliced from four bits, three from the Silene
vulgarise mitochondria chr1 (i.e. NC_016406.1) from both strands, and
one from mitochondria chr3 (NC_016402.1). Just in case there was any
confusion, the human readable note confirms this - not the CDS feature
has a join of five parts, while the gene has a join of just four - there is
an intron too.

Looks like a bug in our parser...

Peter