# How to dump genes from GenBank in GFF3 format?

This question has also been asked on BioStars

If I look at this record in GenBank I see about 6k genes: https://www.ncbi.nlm.nih.gov/nuccore/CM000760?report=gbwithparts

I'd really like to be able to dump those genes in GFF3 format, but I'm guessing I can't do that in a single step, so...

1. How to dump all genes from a GenBank record in 'whatever' format?
2. How to convert 'whatever' format to GFF3?

Turns out you can just grab the GFF3 from the NCBI's FTP site!

ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCA/000/003/195/GCA_000003195.3_Sorghum_bicolor_NCBIv3/

See:

Or "Access the data" on the right here:

• That's useful, but it isn't actually answering the question you asked. Could you please either change the question so it is asking where you can get the data (instead of how to convert GenBank/anything to GFF) or delete either your answer or the question? I know this is kind of weird, but these are strictly Q&A sites, so answers should be answering the question asked. Jun 21 '17 at 15:41

When accessing all of the annotated genes for a reference genome, downloading a GFF3 file directly from the Genbank or RefSeq FTP sites is definitely the way to go.

But when submitting a search query into NCBI Entrez, there is no easy way to download the results in GFF3 format. Probably the best route is to download the results in ASN.1 format, and then use annotwriter from the NCBI C++ toolkit to convert to GFF3. But getting the latter to compile and run properly isn't trivial in my experience.