I have a list of scientific names of species. Is there a python module that can fetch NCBI taxonomy IDs?
$\begingroup$
$\endgroup$
1
-
1$\begingroup$ Biopython has a collection of NCBI Entrez query modules. Please check this StackOverflow thread. stackoverflow.com/questions/16504238/… $\endgroup$– arupSep 5, 2019 at 9:35
Add a comment
|
3 Answers
$\begingroup$
$\endgroup$
This isn't a Python library, but it's easily invoked from Python code, easy to install, and has a lot of useful features.
I just discovered the amazing taxonkit library this week. In particular, the taxonkit name2taxid command seems to be what you're looking for.
$ cat NAMES
Polistes dominula
Sinorhizobium meliloti
Gossypium hirsutum
$ cat NAMES | taxonkit name2taxid
Polistes dominula 743375
Sinorhizobium meliloti 382
Gossypium hirsutum 3635
Easily installed using conda install taxonkit.
$\begingroup$
$\endgroup$
Biopython has an efetch frontend.
from Bio import Entrez
Entrez.email = "mg@example.edu"
handle = Entrez.efetch(db="nucleotide", id="KR052012.1", rettype="gb", retmode="text")
print(handle.read())
OUT
LOCUS KR052012 10714 bp RNA linear VRL 09-MAY-2015
DEFINITION Dengue virus strain Hb33/CHN/2014, complete genome.
ACCESSION KR052012
VERSION KR052012.1
KEYWORDS
.... etc
You will likely need to perform an EFetch search and then extract out the accession numbers for import (as done above)
$\begingroup$
$\endgroup$
If you want to perform a species search, here's the raw efetch code. I use this within an alias and the complex quotes are needed for complicated reasons. The reg-ex may need to be adapted to your search. $virus here is your species of interest, e.g. set virus='some species' (csh or tcsh), `virus='some species' (bash). You'll just need to translate this to Biopython.
alias esummary-acc \
'esearch -db protein -query "$virus" |\
efetch -format docsum |\
xtract -pattern DocumentSummary -element Title,AccessionVersion |\
grep "$virus" |\
perl -p -e '"'"'s/^.*\[.*\]\s+(\w+)/$1/'"'"''