Get taxonomy from genbank accessions

acc2sci

#!/bin/bash

## Fetch scientific name for Genbank accession number
## Version: Mon 20 nov 2017 13:35:49
## By:      {Andreas.Kahari,Johan.Nylander}@nbis.se
## Usage:   ./acc2sci.sh infile-with-accessions
## Notes:   Reads from file or STDIN, prints to STDOUT.
##          No detailed error checking! If any of the Accessions
##          provided as queries are incorrect, the script will
##          exit with msg: "Failure of post to find data to load.."
## Requirements: Entrez Direct suite of software
##       (https://www.ncbi.nlm.nih.gov/books/NBK179288/)

set -euo pipefail

#author_email_address='[email protected]'
#name_of_script='acc2tax'
#econtact -email ${author_email_address} -tool ${name_of_script}

### Andreas' solution
#rm -f __taxid-acc.out __taxonomy.out
xargs -n 200 <${1:-/dev/stdin} | tr ' ' ',' |\
    xargs -n 1 sh -c '
            esearch -db nuccore -query "$0" |\
            efetch -format docsum |\
            xtract -pattern DocumentSummary -element TaxId,Caption |\
            tee -a __taxid-acc.out | cut -f1 | sort -u |\
            epost -db taxonomy |\
            efetch -format docsum |\
            xtract -pattern DocumentSummary -element TaxId,ScientificName,Division' \
>__taxonomy.out

join -t$'\t' \
  -o 1.2,2.1,2.2,2.3 \
  <( sort -u __taxid-acc.out ) <( sort -u __taxonomy.out )
### End of Andreas' solution
rm -f __taxid-acc.out __taxonomy.out

## Slow, check one by one
#while read line
#do
#    sleep 1
#    echo -n "$line "
#    esearch -db nuccore -query "${line}" < /dev/null |\
#    efetch -format docsum |\
#    xtract -pattern DocumentSummary -element TaxId |\
#    sort -n |\
#    uniq |\
#    epost -db taxonomy |\
#    efetch -format docsum |\
#    xtract -pattern DocumentSummary -element ScientificName,Division,TaxId
#done < "${1:-/dev/stdin}"