[EMBOSS] can't access databases indexed by dbifasta

[Jérôme Laroche]

Marcus,

yes, the indexed files will always have the same names (I think this is  
in the EMBOSS administrator's guide):

acnum.hit  acnum.trg  division.lkp  entrynam.idx

That's why you must put these files in different directories. The  
sequence files could be in the same directory.


Jerome

Centre de bioinformatique
Pavillon Charles-Eugène-Marchand
Bureau 4233d
tél: 418-656-2131 poste 6184
fax: 418-656-7176
www.bioinfo.ulaval.ca
Le 04-02-18, à 09:52, Marcus Claesson a écrit :

> That works fine now, thanks!
>
> But I continue struggling with upcoming problems and you must be tired
> of me asking questions all the time... (It's either me or the emboss  
> man
> pages that are bad, probably a mix ;))
>
> Anyway, when I want to create a second database ecoli_aa, dbifasta  
> seems
> to overwrite the old indexfiles.
>
> This works all fine:
>
> [marcus]$ dbifasta -idformat ncbi -directory /var/data/blast_db
> -filenames ecoli.nt -fields des -dbname ecoli_nt -release 0.0 -date
> 00/00/00
> [marcus]$ seqret ecoli_nt: -auto -stdout|head
>> AE000111 AE000111.1 Escherichia coli K-12 MG1655 section 1 of 400 of
> the complete genome
> AGCTTTTCATTCTGACTGCAACGGGCAATATGTCTCTGTGTGGATTAAAAAAAGAGTGTC
> TGATAGCAGCTTCTGAACTGGTTACCTGCCGTGAGTAAATTAAAATTTTATTGACTTAGG
> TCACTAAATACTTTAACCAATATAGGCATAGCGCACAGACAGATAAAAATTACAGAGTAC
> ACAA
>
> I add a second database:
>
> [marcus]$ dbifasta -idformat ncbi -directory /var/data/blast_db
> -filenames ecoli.aa -fields des -dbname ecoli_aa -release 0.0 -date
> 00/00/00
>
> And when I seqret the old ecoli_nt I get the new ecoli_aa entries
> instead:
>
> [marcus]$ seqret ecoli_nt: -auto -stdout|head
>> AAC73112.1 (AE000111) thr operon leader peptide [Escherichia coli]
> MKRISTTITTTITITTGNGAG
>> AAC73113.1 (AE000111) aspartokinase I, homoserine dehydrogenase I
> [Escherichia coli]
> MRVLKFGGTSVANAERFLRVADILESNARQGQVATVLSAPAKITNHLVAMIEKTISGQDA
> LPNISDAERIFAELLTGLAAAQPGFPLAQLKTFVDQEFAQIKHVLHGISLLGQCPDSINA
> ALICRGEKMSIAIMAGV
>
>
> I tried to run dbifasta for both fasta files at the same time but it
> didn't work:
> dbifasta -idformat ncbi -directory /var/data/blast_db -filenames
> ecoli.nt ecoli.aa -fields des -dbname ecoli_nt ecoli_aa -release 0.0
> -date 00/00/00
>
> How can new databases be added without overwriting the old indexfiles?
>
>
> /Marcus
>
>
>
>
> On Wed, 2004-02-18 at 13:50, Gary Williams, Tel 01223 494522 wrote:
>> "Gary Williams, Tel 01223 494522" wrote:
>>>
>>> Marcus Claesson wrote:
>>>> But this doesn't work and I don't know why since I added the 'des
>>>> field'. Are queries like these suppose to work by the way?
>>>>
>>>> [marcus]$ seqret ecoli_nt:\*mitochondrio\* -auto -stdout
>>>> Error: Unable to read sequence 'ecoli_nt:*mitochondrio*'
>>>> Died: seqret terminated: Bad value for '-sequence' with -auto  
>>>> defined
>>>
>>> The correct syntax should be:
>>> seqret ecoli_nt-des:\*mitochondrio\* -auto -stdout
>>> See:
>>> http://www.rfcgr.mrc.ac.uk/Software/EMBOSS/Themes/ 
>>> UniformSequenceAddress.html#keys
>>
>> I think that the '*' are not required, so:
>> seqret ecoli_nt-des:mitochondrio -auto -stdout
>