Download the GFF file of the Cycas taitungensis mitochondrion

[sjackman]

Hi. I'd like to download the annotation in GFF format of the Cycas taitungensis mitochondrion, NC_010303.1. I started with bionode ncbi search genome cycas taitungensis, but got nada back. Any suggestions?

[bmpvieira]

Hi, there's no ncbi "genome" or "assembly for Cycas taitungensis, but you can download the Genbank file and convert it to GFF, or just use togows.org to get it directly using the NCBI ID:

bionode-ncbi search nuccore NC_010303.1 | json uid | while read id; do wget http://togows.org/entry/nucleotide/$id.gff; done

[sjackman]

Perfect! Thanks, Bruno.

http://togows.org/entry/nucleotide/166895601.gff

[sjackman]

For future me, or someone else, here's my Makefile snippet:

# Cycas taitungensis
NC_010303.1.json: %.json:
    bionode-ncbi search nuccore $* >$@

%.uid: %.json
    json uid <$< >$@

%.fa: %.uid
    curl http://togows.org/entry/nucleotide/`<$<`.fasta |seqtk seq >$@

%.gb: %.uid
    curl -o $@ http://togows.org/entry/nucleotide/`<$<`.gb

%.gff: %.uid
    curl -o $@ http://togows.org/entry/nucleotide/`<$<`.gff

[sjackman]

npm install -g bionode-ncbi json

[bmpvieira]

Thanks @sjackman for your comments. I haven't closed this yet because it could be a nice thing to add to bionode. /cc @terfn

[bmpvieira]

@terfn recently added NCBI efetch to bionode, so soon something like this should work:

echo -e "{'db': 'nuccore', 'term': 'NC_010303.1', 'rettype': 'fasta'}\n" | bionode-ncbi fetch -

It currently doesn't because of the way CLI is parsing things, but in JavaScript the following works and gives a fasta:

var ncbi = require('bionode-ncbi')
var fasta = require('bionode-fasta')
var json = require('JSONStream')

var opts = {
  db: 'nuccore',
  term: 'NC_010303.1',
  rettype: 'fasta'
}

ncbi.fetch(opts)
.pipe(json.stringify()) // We shouldn't need this, bionode-fasta should accept objects
.pipe(fasta.write())
.on('data', function (data) {
  console.log(data)
})

[terfilip]

By the way fasta is the default rettype for nuccore and some other dbs so it's not necessary to specify it. Also it is possible to get the GenBank flat file format from nuccore using efetch by specifying

rettype: 'gbwithparts'

If the goal is to convert it to a GFF later

@bmpvieira I'll take a look at making the CLI accept objects

[bmpvieira]

After #32 is merged, this should work:

echo -e '{"term": "NC_010303.1", "rettype": "fasta"}' | bionode-ncbi fetch nuccore -s

[sjackman]

Nice. Thanks, Bruno!