Re: FW: [GUSDEV] Load a Blast result using LoadNRDB

[Deborah P. <pi...@pc...>]

And finally I see your attachment and it looks like it is in the right=20
format to be loaded with LoadNRDB. However, if the defline only has a=20
single source_id, I think I would use LoadFastaSequences.pm to load your=20
sequences into dots.AASequence or one of its views. This plugin is=20
generic and loads your choice of sequence tables with source_id, etc=20
defined using regex. I think you will find that more satisfactory.

                                                                         =
                                                                        =20
Debbie




Deborah Pinney wrote:

> I reread your e-mail and am guessing that your file is a file of=20
> sequences (query or subject ?) with information on the defline=20
> pertaining to a BLAST analysis. Perhaps you can send an excerpt of=20
> your file  and a bit more explanation. Perhaps we can make some better=20
> suggestions.
>
>                                                                        =
                                    =20
> Debbie
>
>
>
> Deborah Pinney wrote:
>
>> LoadNRDB.pm is not the appropriate plugin to use to load BLAST=20
>> results into GUS. It is specifically designed to load defline and=20
>> sequence information  from NCBI's non redundant protein database=20
>> files. It uses the gitax file to get taxon information for each of=20
>> the gi numbers in the defline so as to attach a taxon_id. Therefore,=20
>> LoadTaxon would have to be run with NCBI's taxonomy files prior to=20
>> running LoadNRDB.
>>
>> There is at least one plugin that can be used to load BLAST results,=20
>> InsertBlastSimilarities.pm, which is one of the GUS/Supporterd=20
>> plugins. Take a look at that and see if it serves your purposes.
>>                                                                       =
                                                                         =
    =20
>> Debbie
>>
>>                                                =20
>>
>>
>> Michael Saffitz wrote:
>>
>>> Hi Debbie,
>>>
>>> Can you help with this?  I think we should start with an overview of=20
>>> what
>>> LoadNRDB is doing-- specifically how it handles the gitax file.  We=20
>>> can then
>>> figure out if Fabricio's issue is in the data or code.
>>>
>>> Please include gusdev in your reply-- I think it will be useful=20
>>> information.
>>>
>>> --Mike
>>>
>>> ------ Forwarded Message
>>> =20
>>>
>>>> From: Fabr=EDcio <fab...@de...>
>>>> Date: Thu, 9 Jun 2005 16:38:44 -0300
>>>> To: <gus...@li...>
>>>> Subject: [GUSDEV] Load a Blast result using LoadNRDB
>>>>
>>>> Hello all,
>>>>
>>>>
>>>> We=B9re trying to insert a Blast result into Gus Schema using LoadNR=
DB
>>>> plug-in. Our blast result was filtered and is in a Fasta format=20
>>>> according to
>>>> the plug-in requirement. The problem is the load process takes many=20
>>>> times
>>>> and our postgres server process crashes after some hours, thus the=20
>>>> plug-in
>>>> never concludes. I would like to know why this plug-in is so slow,=20
>>>> if the
>>>> file we want to store is not so big. I think that it=B9s due to the
>>>> =ADgitax=3Dgi_taxid_prot.dmp parameter that has a large file.
>>>>
>>>>
>>>>
>>>> Does anyone could help us to understand this?
>>>>
>>>>
>>>>
>>>> Thanks a lot,
>>>>
>>>>
>>>>
>>>> Fabr=EDcio.
>>>>
>>>>  =20
>>>
>>>
>>>
>>> ------ End of Forwarded Message
>>> =20
>>>
>>
>>
>>
>> -------------------------------------------------------
>> This SF.Net email is sponsored by the 'Do More With Dual!' webinar=20
>> happening
>> July 14 at 8am PDT/11am EDT. We invite you to explore the latest in du=
al
>> core and dual graphics technology at this free one hour event hosted=20
>> by HP, AMD, and NVIDIA.  To register visit=20
>> http://www.hp.com/go/dualwebinar
>> _______________________________________________
>> Gusdev-gusdev mailing list
>> Gus...@li...
>> https://lists.sourceforge.net/lists/listinfo/gusdev-gusdev
>
>