Species-Specific Output From Blast -Db Nt -Remote ...

Question

[kevin.l.neff]

I'm using the C++ version of blastn, with -db nt and -remote. I need to limit my search to a specific organism. I've had success using -gilist <file> to do this on searches with a local database, but that option is excluded for remote searches. Should I use a script to filter the search results, like -gilist, or is there an easier way?

[Lee Katz]

I would also like to know this, but for local blast vs nr!

Answer 1

[chris.mit7]

Add -entrez_query "Homo Sapiens[Organism]" to your search or whatever organism you want.

Such as:

blastn -remote -db nr -query stuff_to_search.fasta -entrez_query "Homo Sapiens[Organism]" -evalue 1e-20 -num_alignments 10 > stuff_to_search.blastn

[Kumar]

How about if more than one species.

Thanks

Answer 2

[Paulo Nuin]

Use the -gilist option in the command line. Create a file with the GIs you want, for instance all bacterias, all mammals, one frog, one lizard, etc and call that file using the option

... -gilist prots.txt ...

[kevin.l.neff]

Sadly, this doesn't work with -remote option.

[Paulo Nuin]

Yes, you are right. IMHO, you would be best to download the nr DB and run locally.

[Lee Katz]

How would you generate this gi list for an entire genus or other taxonomy?

[Paulo Nuin]

Best option would be to go to NCBI website and browse via Taxonomy. This way it is possible to go down/up the tree and check the GIs, as mostly there's one set be genus, species, etc.

[kevin.l.neff]

Yes, I agree. However, my web server only allows 10 GB data, so I need to come up with an alternative. Dreamhost or some other service may be the easiest solution, but I wondered if there was a way to do this without bringing in the big guns. Seems like no...

[kevin.l.neff]

http://www.biostars.org/post/show/15002/make-a-custom-blast-library-using-the-output-of-another-blast-result/#45602