protein gi number to description

[mkdir]

Hello,

I got gi numbers for blastp -outfmt 6 results. I am wondering which programs and database can convert the gi number list to protein descriptions. Any help? Thank you very much.

[maubp]

In theory BLAST could do this for you from the description in the database, please email them to let them know people would use this:

http://blastedbio.blogspot.co.uk/2012/05/blast-tabular-missing-descriptions.html

But as to the general question, you want to go from a protein GI number, e.g. GI:7525018, to a description, here something like "ATP synthase CF1 alpha subunit"?

You could use the NCBI Entrez API to lookup by protein GI number, see http://www.ncbi.nlm.nih.gov/books/NBK25499/

Another online alternative is the TogoWS REST API http://togows.dbcls.jp/site/en/ which makes this very easy thanks to their option to extract a specific field from a record, and can be used at the command line, in a script, or in your web-browser:

$ curl [url]http://togows.dbcls.jp/entry/protein/7525018/definition[/url]
ATP synthase CF1 alpha subunit (chloroplast) [Arabidopsis thaliana].

[kmcarr]

Hello,

I got gi numbers for blastp -outfmt 6 results. I am wondering which programs and database can convert the gi number list to protein descriptions. Any help? Thank you very much.

I'm guessing from the fact that you said you used '-outfmt 6' that you ran your BLAST search locally, using BLAST+, with a local database. If that is the case then you already have the toolset and data needed. You can use the blastdbcmd which is part of the BLAST+ distribution. This command should give you what your want.

# blastdbcmd -db <your_blastdb_name> -entry_batch <gi_list_file> -outfmt '%g %t' -target_only -out <output_file_name>

The gi_list_file contains the gi's you want to search, one entry per line. The outfmt will print "gi#<space>definition", one per line. You can adjust the outfmt as desired (blastdbcmd -help for a full description of possible formats).

[maubp]

That's rather neat kmcarr, in this case for just the one GI number in my example:

$ blastdbcmd -db /data/blastdb/ncbi/nr -entry 7525018 -outfmt '%t' -target_only
ATP synthase CF1 alpha subunit [Arabidopsis thaliana]

Nice

[mkdir]

Thank you very much kmcarr and maubp. This worked out.