[Bioperl-l] Constructing split features

[Warren Gallin]

It occurs to me that if one could create a Bio::SeqFeature object using the information in a coded_by tag from a CDS Bio::SeqFeature in a protein record, then it should be possible to add that new feature to a copy of the underlying nucleotide sequence (just pull the sequence from GenBank in fast format, create a temporary new GenBank format Bio::Seq object, add the single new feature and use it to extract the spliced sequence).

Is this even possible?  Or would you have to strip out all the information on the various spliced locations from the coded-by tag and rebuild a new Bio::SeqFeature from scratch?  Or is this even a reasonable approach to creating an automated way of getting coding sequences?

Warren Gallin