How to get taxonomic specific ids for kingdom, phy

问题:

I have a list of taxids that looks like this:

1204725
2162
1300163
420247

I am looking to get a file with taxonomic ids in order from the taxids above:

kingdom_id      phylum_id       class_id        order_id        family_id       genus_id        species_id   

I am using the package "ete3". I use the tool ete-ncbiquery that tells you the lineage from the ids above. (I run it from my linux laptop with the command below)

ete3 ncbiquery --search 1204725 2162 13000163 420247 --info 

The result looks like this:

# Taxid Sci.Name    Rank    Named Lineage   Taxid Lineage
2162    Methanobacterium formicicum species root,cellular organisms,Archaea,Euryarchaeota,Methanobacteria,Methanobacteriales,Methanobacteriaceae,Methanobacterium,Methanobacterium formicicum   1,131567,2157,28890,183925,2158,2159,2160,2162
1204725 Methanobacterium formicicum DSM 3637    no rank root,cellular organisms,Archaea,Euryarchaeota,Methanobacteria,Methanobacteriales,Methanobacteriaceae,Methanobacterium,Methanobacterium formicicum,Methanobacterium formicicum DSM 3637  1,131567,2157,28890,183925,2158,2159,2160,2162,1204725
420247  Methanobrevibacter smithii ATCC 35061   no rank root,cellular organisms,Archaea,Euryarchaeota,Methanobacteria,Methanobacteriales,Methanobacteriaceae,Methanobrevibacter,Methanobrevibacter smithii,Methanobrevibacter smithii ATCC 350611,131567,2157,28890,183925,2158,2159,2172,2173,420247

I have no idea which items (IDS) correspond to what I am looking for (if any)

回答1:

The following code:

import csv
from ete3 import NCBITaxa

ncbi = NCBITaxa()

def get_desired_ranks(taxid, desired_ranks):
    lineage = ncbi.get_lineage(taxid)
    lineage2ranks = ncbi.get_rank(lineage)
    ranks2lineage = dict((rank, taxid) for (taxid, rank) in lineage2ranks.items())
    return {'{}_id'.format(rank): ranks2lineage.get(rank, '<not present>') for rank in desired_ranks}

def main(taxids, desired_ranks, path):
    with open(path, 'w') as csvfile:
        fieldnames = ['{}_id'.format(rank) for rank in desired_ranks]
        writer = csv.DictWriter(csvfile, delimiter='\t', fieldnames=fieldnames)
        writer.writeheader()
        for taxid in taxids:
            writer.writerow(get_desired_ranks(taxid, desired_ranks))

if __name__ == '__main__':
    taxids = [1204725, 2162,  1300163, 420247]
    desired_ranks = ['kingdom', 'phylum', 'class', 'order', 'family', 'genus', 'species']
    path = 'taxids.csv'
    main(taxids, desired_ranks, path)

Produces a file that looks like this:

kingdom_id  phylum_id   class_id    order_id    family_id   genus_id    species_id
<not present>   28890   183925  2158    2159    2160    2162
<not present>   28890   183925  2158    2159    2160    2162
<not present>   28890   183925  2158    2159    2160    2162
<not present>   28890   183925  2158    2159    2172    2173

回答2:

With the Taxid Lineage numbers in your results, try using them in ete3's get_rank method. As an example:

from ete3 import NCBITaxa
ncbi = NCBITaxa()

print ncbi.get_rank([9606, 9443])
# {9443: u'order', 9606: u'species'}

Presumably the resulting dictionary should contain the rank information of all IDs, including any intermediate "no rank" IDs that you may want to eliminate.

回答3:

You can also use the R packaage taxonomizr. The package takes a bit of time to download the necessary files, but after that its quite fast and easy.

library("taxonomizr)
getNamesAndNodes()
taxaNodes <- read.nodes('nodes.dmp')
taxaNames <- read.names('names.dmp')
taxaID <- c("1204725", "2162", "1300163", "420247")

getNamesAndNodes downloads the names.dmp and nodes.dmp file from ncbi.