This is a test version of Biostars. For the public version, visit https://www.biostars.org.
how to makeblastdb with taxon id's

Dear all,

I extracted a subset of sequences from the nt database and I want to make a blast database with those sequences including the taxon id's.

If I execute the following command :

./ncbi-blast-2.6.0+/bin/makeblastdb -in ntselection.fa -dbtype nucl -taxid_map nucl_gb.accession2taxid -parse_seqids

I get no errors and the database is working with blast, but with output parameter -outfmt '6 qseqid sseqid stitle sgi sacc pident length qlen evalue bitscore staxids' there are no taxon id's in the output.

What is the correct command and input files? If I use the gi_taxid_nucl.dmp I also get no errors but no blast database is made.

makeblastdb blast
• 5,656 views • link
updated by Biostar Biostar • written by gb gb

The taxdb is to retrieve the scientific name from the taxid. I need the taxid in my blast output. In the last column I now only see 'N/A'. But I do know that there are taxid's available because I can find them in the file nucl_gb.accession2taxid (ftp://ftp.ncbi.nih.gov/pub/taxonomy/accession2taxid/)

• 0 views • link
written by gb gb

I am aware of that post and already using blast 2.6.0 and it is still not working. I do not get errors, if I use the same format taxon map file as in that post and use blast 2.6.0 I do not get errors but there is also no database made. I think I go try blast 2.3.0 with one of the solutions

• 0 views • link
written by gb gb

If you use the search, you'll find out that this has been asked before and at least back then there was no direct solution. However, post blast it's very easy to add that information to your output. Just use join and sort, e.g.:

join -t $'\t' -1 1 -2 1 -o 1.1,1.2,1.3,...,2.2 \
    <(sort -t $'\t' -k1,1 blastoutput) \
    <(sort -t $'\t' -k1,1 nucl_gb.accession2taxid)
• 0 views • link
written by 5heikki 5heikki

It sounded that upgrading to 2.6.0 was the solution. If I want to use your solution I have to adjust someone else's pipeline and was trying to avoid that. According the blast documentation it should be possible.

Anyways thanks for the answer

• 0 views • link
written by gb gb

I have some weird results now, I use one sequence to make the blast database for test purposes.

If I use an accession_taxonid file consisting of one line (the accession and taxon id of that single sequence) it works!

But, if I use the same makeblastdb command with the complete accession_taxonid file there is no database made... Maybe the length and order of the accession_taxonid file must be the same as the input sequences.

• 0 views • link
written by gb gb

0 answers

No answers yet.

Log in to answer this question.

More posts like this