[Bioperl-l] Bio::SeqFeatureProducer

Jason Stajich jason@chg.mc.duke.edu
Thu, 14 Dec 2000 17:59:36 -0500 (EST)


(Hilmar, if you can help be my verifier that we have completed all the
parts of the task, we can start to tick off some of the things from the
tasklist...)

I commited a new class Bio::SeqFeatureProducer, a test file, and more test
data.  

One can use this class to simplify the way Features from an analysis are
added to a sequence (Genes from Bio::Tools::Genscan or Bio::Tools::MZEF).  

In some cases you will want more fine grained control over how features 
are added or seen so that more information is produced in a Genbank/EMBL
dump.  This can be done by cycling through the features on the sequence
and processing them.  ie I want to dump CDS and predicted protein
translation in genbank files from genscan analysis, I have to take each
gene on the seq, add a new feature for CDS with a translated field.
Perhaps we can work out a way for the GeneStruct objects to be dumped a
little more informatively?

I suspect we will want to remove Bio::SeqFeatureProducerI as I don't
think it is necessary unless we will have many different ways to
add_features to sequences with non AnalysisParserI objects...  

Anyone else writing parsers - GAME w/o seq and gff will certianly fit here
with a simple wrapper - should make sure they have a class that implements
SeqAnalysisParserI and then SeqFeatureProducer can take advantage of the
class.  Will have to add them to the %DRIVERVALUES variable in
SeqFeatureProducer since we don't have parsers in a central location.

Hope this is useful.

-jason

Jason Stajich
jason@chg.mc.duke.edu
Center for Human Genetics
Duke University Medical Center 
http://www.chg.duke.edu/