next up previous contents
Next: DIRECT Up: Database access methods Previous: APP   Contents

BLAST

Modes:

a q s

BLAST uses EMBLCD indices created with dbiblast to access databases in BLAST format, created with NCBI's

formatdb
program.

Note that the latest 'format version 4' is not yet documented by NCBI. EMBOSS will only work with 'format version 3' databases, indexed with:

formatdb -A F

We hope to support 'format version 4' databases in future. If you pick up a blast database from NCBI (or elsewhere) check the format. If it is in the new format, you will need to pick up the original FASTA format file, and either index it yourself with formatdb, or run

dbifasta
and use the FASTA file in EMBOSS (see EMBLCD access method)

The definition should use format: ncbi because this is what the blast formatdb databases store internally.

DB mydb [
#required parameters
   method: "blast"
   format: "ncbi"
   type: "N"
   dir: "\$emboss_db_dir/blas"t
#optional parameters
   fields: "sv des"
   release: "63.0"
   comment: "my comment"
   indexdir: "\$emboss_db_dir/blastindices"]

The index files can be kept in the same directory as the database, but as each EMBLCD index needs its own directory (the filenames are fixed) the indexdir is usually defined.

The EMBLCD index files include the filenames indexed by

dbiblast
. You can use the file: and exclude: attributes to create file-specific subsets from a single
dbiblast
generated index, but as blast index files are split only by the number of entries this is not generally useful.

If the database was indexed with additional fields, they can be included in the definition as fields: to allow their use in USAs.


next up previous contents
Next: DIRECT Up: Database access methods Previous: APP   Contents
Peter Rice 2007-04-26