[EMBOSS] dbifasta and ncbi formats

[pmr at ebi.ac.uk]

Hi Tonu,

> If I run dbifasta and choose –idformat ncbi  then after creating database
> index seqret returns sequences only when I specify NP_414542.1 and not by
> gi
> number.  How can I use the gi number for retrival as well?

Dbifasta indexes only ID and accession by default.

To get the other index fields you can specify -fields "des,acc,sv" when
you run dbifasta, and add fields: "des sv" to the database definition. You
can then search by GI and/or sequence version, if EMBOSS can find them in
the NCBI FASTA format your data is in.

We do not recommend dbiblast - we are unable to decode all the index file
formats formatdb can produce, so we only can support an old version of
formatdb and not the files currently provided by NCBI's FTP server). We
have decoded about 50% of the file so far, but it is a difficult job with
limited documentation to decode the remainder. If anyone does have
documentation we would be happy to take another look.

regards,

Peter