[Bioperl-l] BLASTing BioSeq objects

James Wasmuth james.wasmuth@ed.ac.uk
Tue, 10 Dec 2002 17:35:47 +0000


Sorry to fill every1's intrays, but this has bugged me for a while and 
has caused me to miss my bus  :o(

rigtheo, I manged to get this far thru a few other methods, but Shawn's 
was the most concise.

So my very rough code looks like...

sub CREATEBIOSEQ    {
    my $query_str = $_[0];          #query sequence in fasta
   my $stringfh = new IO::String($query_str);
    my $sio = new Bio::SeqIO(-fh => $stringfh, -format => 'fasta');
    my $bioSeq_obj = $sio->next_seq;  # create Bio::Seq object
    return $bioSeq_obj;
}

$bioSeq_obj has size (checked it using length).  However when I place it 
into Blast later I get...

MSG:  Bio::Seq=HASH(0x8620500) (0) not Bio::Seq object or array of 
Bio::Seq objects or file name!

Have tried all the checks I can and the object is the same one all the 
way thru the process, but something isn't recognising it!
If its nothing obvious then I'll have a hunt round the rest of my code 
and see if it comes from there...

Thanx
J



Shawn wrote:

>actually, $sio->next_seq is a handle that returns a Bio::Seq object.
>so u can pass what it returns directly to blast.
>
>my $fact = Bio::Tools::Run::StandAloneBlast->new('program'=>'blastn',
>	                                        'database'=>"mydb",
>		                               'outfile'=>'tmp/xxx');
>
>
>my $stringfh = new IO::String($query_str);
>my $sio = new Bio::SeqIO(-fh => $stringfh, -format => 'fasta');
>$fact->blastall($sio->next_seq);
>
>
>
>On Wed, 2002-12-11 at 01:08, James Wasmuth wrote:
>  
>
>>That would give me a SeqIO object, Blast requires a Bio::Seq object.  So 
>>I would then need to convert $seq to a Bio::Seq object?  If so, this is 
>>fine, and thanks.
>>James
>>
>>Shawn wrote:
>>
>>    
>>
>>>oh ok, misread you. Your original solution was almost there.
>>>this works for me:
>>>
>>>use Bio::SeqIO;
>>>use IO::String;
>>>my $query_str = ">A\nACCCCCCC";
>>>      
>>>
>
>  
>
>>>my $stringfh = new IO::String($query_str);
>>>my $sio = new Bio::SeqIO(-fh => $stringfh, -format => 'fasta');
>>>my $seq = $sio->next_seq;
>>>
>>>print $seq->seq;
>>>
>>>
>>>shawn
>>>
>>>On Wed, 2002-12-11 at 00:38, James Wasmuth wrote:
>>> 
>>>
>>>      
>>>
>>>>>my $seq = new Bio::Seq(-seq=>$query_str,-id=>"blast_seq");
>>>>>my $fact = Bio::Tools::Run::StandAloneBlast->new('program'=>'blastn',
>>>>>	                                        'database'=>"mydb",
>>>>>		                               'outfile'=>'tmp/xxx');
>>>>>my $blast_report = $fact->blastall($seq);
>>>>>
>>>>>
>>>>>     
>>>>>
>>>>>          
>>>>>
>>>>Unfortunately I've tried something along these lines.  The problem seems 
>>>>to caused by the query sequence being in fasta format.  In the error log 
>>>>I get a message...
>>>>
>>>>"MSG: Attempting to set the sequence to [>GLBH_CAEEL
>>>><--some aa-->] which does not look healthy"
>>>>
>>>>I know that this is because entering a sequence using '-seq' it has to 
>>>>be raw and so have no fasta header or "\n" characters, so I am trying to 
>>>>get round this.  I've done it by storing the query sequence as a file 
>>>>and using this to blast with but had been hoping to avoid this...
>>>>
>>>>James
>>>>
>>>>
>>>>
>>>>Shawn wrote:
>>>>
>>>>   
>>>>
>>>>        
>>>>
>>>>>You probably want to do:
>>>>>
>>>>>
>>>>>cheers,
>>>>>
>>>>>shawn
>>>>>
>>>>>On Tue, 2002-12-10 at 23:40, James Wasmuth wrote:
>>>>>
>>>>>
>>>>>     
>>>>>
>>>>>          
>>>>>
>>>>>>Hi all,
>>>>>>
>>>>>>I'm trying to blast a sequence which is introduced to the program as a 
>>>>>>string all implemented as a CGI userinterface.  The sequence is in fasta 
>>>>>>format, though I hope to extend this to any.
>>>>>>
>>>>>>I thought I had been able to create a Bio::Seq object from the string, 
>>>>>>which StandAloneBlast requires,  by doing...
>>>>>>
>>>>>>my $stringfh = new IO::String($query_str);
>>>>>>my $bioSeq_obj = new Bio::Seq(-fh => $stringfh, -format => 'fasta');
>>>>>>
>>>>>>however a check on the length of the sequence in the object reveals it 
>>>>>>to be zero in length.
>>>>>>
>>>>>>Anyone any ideas?  Should I first create a SeqIO object and convert this 
>>>>>>to a Bio::Seq object and then BLAST.  If so, how is this conversion done?
>>>>>>
>>>>>>I'm sure the answers are in the archive but have been unable to locate 
>>>>>>them...
>>>>>>
>>>>>>Many Thanks
>>>>>>James
>>>>>>
>>>>>>_______________________________________________
>>>>>>Bioperl-l mailing list
>>>>>>Bioperl-l@bioperl.org
>>>>>>http://bioperl.org/mailman/listinfo/bioperl-l
>>>>>>
>>>>>>  
>>>>>>
>>>>>>       
>>>>>>
>>>>>>            
>>>>>>
>>>>>     
>>>>>
>>>>>          
>>>>>
>>>>-- 
>>>>
>>>>Nematode Bioinformatics
>>>>Blaxter Nematode Genomics Group
>>>>Institute of Cell, Animal and Population Biology
>>>>Ashworth Labs
>>>>University of Edinburgh
>>>>King's Buildings
>>>>Edinburgh
>>>>EH9 3JT
>>>>
>>>>0131 650 7403
>>>>
>>>>
>>>>
>>>>_______________________________________________
>>>>Bioperl-l mailing list
>>>>Bioperl-l@bioperl.org
>>>>http://bioperl.org/mailman/listinfo/bioperl-l
>>>>
>>>>   
>>>>
>>>>        
>>>>
>>> 
>>>
>>>      
>>>
>>-- 
>>
>>Nematode Bioinformatics
>>Blaxter Nematode Genomics Group
>>Institute of Cell, Animal and Population Biology
>>Ashworth Labs
>>University of Edinburgh
>>King's Buildings
>>Edinburgh
>>EH9 3JT
>>
>>0131 650 7403
>>
>>
>>
>>    
>>
>
>
>  
>

-- 

Nematode Bioinformatics
Blaxter Nematode Genomics Group
Institute of Cell, Animal and Population Biology
Ashworth Labs
University of Edinburgh
King's Buildings
Edinburgh
EH9 3JT

0131 650 7403