[Bioperl-l] Problem with parsing ENSEMBL genbank flat file wi
th genbank2gff3. pls
Babenko, Vladimir (NIH/NLM/NCBI)
babenko at ncbi.nlm.nih.gov
Tue Jan 18 14:20:11 EST 2005
Sorry Ewan,
Now I got that when I check multiple transcripts it means genes with no
less than 2 transcripts.
The Mart is amazing.
Regards,
Vladimir
>-----Original Message-----
>From: Ewan Birney [mailto:birney at ebi.ac.uk]
>Sent: Tuesday, January 18, 2005 1:35 PM
>To: Babenko, Vladimir (NIH/NLM/NCBI)
>Cc: 'cjm at fruitfly.org'; 'cain at cshl.edu'; 'jason.stajich at duke.edu'
>Subject: RE: [Bioperl-l] Problem with parsing ENSEMBL genbank
>flat file wi th genbank2gff3. pls
>
>On Tue, 18 Jan 2005, Babenko, Vladimir (NIH/NLM/NCBI) wrote:
>
>> Hi all,
>> Thank you for your prompt response:
>> I believe that gff* is one of the ways to manage data, so
>this is the
>> reason I'm up to this.
>> I played around for a while with all you propositions,
>so further
>> is a short response:
>> 1) Chris - I checked the script, it works fine, thank you. I'm
>> currently exploring this option.
>> The point is that I do need to have both mRNA and CDS linked
>to check
>> for UTR, introns, and looks like genbank2gff3 works fine here.
>> 2) Ensmart - this is a great proposition. I haven't come to
>the end of
>> the investigation of this sound product, but it is my sneaky
>suspicion
>> that I need some kind of mysql dump to manage the stuff by myself in
>> the same way bioperl does, but again it's a compromise between
>> compexity and simplicity that I cannot fully embrace for a while.
>> Still the option of species comparison may annihilate my suspicions
>> momentarily if I will be able to manage it.
>> BTW, Ewan, 'unchecking' for the entire genome and setting multiple
>> transcripts for human yields (all other options unchecked) after
>> Filters stage yields:
>> 7185 Entries pass Filters - that looks a bit few for human.
>Probably I
>> miss out something, sorry.
>
>You want to uncheck the "multiple transcripts" (this means
>genes with more than one transcript: you want all genes).
>
>
>
More information about the Bioperl-l
mailing list