Blastn and Multiple Databases: how best to manage

Question

[jeremy.cox.2]

How do I best manage multiple BLAST databases?

So I am pretty new to Bioinformatics, but I am a computer guy. I have a few questions about how to best use blastn to achieve my goal.

I have multiple databases I prepared, creatively named virus.fa, bacteria.fa, fungi.fa, human.fa, mouse.fa, and rat.fa.

I want to be able to BLAST against any combination of the databases, hopefully without doing anything crazy like computing all database permutations.

1. I don't think I can BLAST against multiple databases at once, is that correct?
2. As I understand it, I can make 6 separate blast databases and blast against them one at a time. Then I concatenate the results.
   1. Is this computationally wasteful?
3. I could make 1 big database and ignore hits for organisms I don't want to search.
   1. This is obviously wasteful.
   2. Can makeblastdb take multiple input files? I don't think it can, so I would have to cat them up before making the db.
4. Is there a solution I am missing, hopefully an elegant solution?

Thank you,
Jeremy Cox
CSE PhD student

Answer 1

[5heikki]

You should just combine the dbs (are those dbs or just fasta files?) and then make aliases for subset dbs..

 blastdb_aliastool -h

USAGE
  blastdb_aliastool [-h] [-help] [-gi_file_in input_file]
    [-gi_file_out output_file] [-db dbname] [-dbtype molecule_type]
    [-title database_title] [-gilist input_file] [-out database_name]
    [-dblist database_names] [-dblist_file file_name]
    [-num_volumes positive_integer] [-logfile File_Name] [-version]

DESCRIPTION
   Application to create BLAST database aliases, version 2.2.29+

   This application has three modes of operation:

   1) GI file conversion:
      Converts a text file containing GIs (one per line) to a more efficient
      binary format. This can be provided as an argument to the -gilist option
      of the BLAST search command line binaries or to the -gilist option of
      this program to create an alias file for a BLAST database (see below).

   2) Alias file creation (restricting with GI List):
      Creates an alias for a BLAST database and a GI list which restricts this
      database. This is useful if one often searches a subset of a database
      (e.g., based on organism or a curated list). The alias file makes the
      search appear as if one were searching a regular BLAST database rather
      than the subset of one.

   3) Alias file creation (aggregating BLAST databases):
      Creates an alias for multiple BLAST databases. All databases must be of
      the same molecule type (no validation is done). The relevant options are
      -dblist and -num_volumes.

[jeremy.cox.2]

Based on the information you provided, it looks like I can keep the 6 databases separate and create an alias file to refer to multiple databases.

blastdb_aliastool -db microbiome -dblist "virus fungi bacteria"

I think this is the opposite of what you described in making alias for subsets?

[5heikki]

Yes, you can do that too..

Answer 2

[jeremy.cox.2]

I found this topic providing helpful answers

How To Blast A Sequence Against Multiple Databases

Sorry for duplicate question