[Bioperl-l] Bug # 2908: ecoli.nt not available

[Iftekharul Haque]

On Sun, Apr 25, 2010 at 9:47 PM, Chris Fields <cjfields at illinois.edu> wrote:
[snip]
>>> Second: I've looked at bug # 2908
>>> http://bugzilla.open-bio.org/show_bug.cgi?id=2908 (ecoli.nt not
>>> available from ftp://ncbi.nlm.nih.gov, causes
>>> t/Tools/Run/StandaloneBlast.t to fail)
>> [snip]
>>> Wondering aloud here: would it make sense to include a compressed
>>> reference file to test against rather than depend on a changing
>>> source? Then this problem could be solved once and for all.
>>
>> Agreed.
>>
>> Actually, we could make the database ourselves and have it contain only a handful of sequences. From my reading of the testfile, there's no need to use a downloaded database to test against here.

Hi Dave,

If we were to include a database ourselves, where would it reside? And
I imagine you use the word "database" fairly loosely, as in, it could
be just a flat text file?

I was thinking if other tools need reference sequences to run tests as
well, if we had a standing set of sequences to test tools against, you
wouldn't have to add too many sequence files with the distribution
(helping control download file size).

With the cornucopia of conversion tools, I imagine a test for any tool
that uses a specific format could easily be coded for in the test
itself (rather than having to include another reference sequence file
for that tool).

I'm not too familiar with the existing corpus of work to tell if this
would be a need looking forward, though.

>>
>>
>> Dave
>
> How does this work cross-platform, cross-OS, 32 vs 64 bit, new vs old BLAST versions, etc?  That's my only concern with using a single database.

Hi Chris,

Do you mean how multi-platform IO would occur with this single database?

Regards,
Ifty.

>
> I don't think there is an issue, just can't recall.
>
> chris
>
>
>
>