[BioPython] ClustalW Multiple Alignment

[David Moreira]

Dear Stefanie,

For a similar case, what I do is to use first BLAST to know the 
orientation of the sequence. I have a small sequence data base with 
sequences in the good orientation and I BLAST against it, BLAST tells 
you whether the retrieved result is "plus/plus" or "minus/plus", i.e., 
whether your query was in the same or different orientation. You have 
just to parse the BLAST output to retrieve that information. BLAST is 
extremely rapid, so you can retrieve the orientation of hundreds of 
sequences in a few minutes. Then, you can reverse-complement the 
sequences in "minus" orientation and construct your multiple sequence 
alignment. It is very easy to have a single script doing all the work.

David

Stefanie Lu"ck a e'crit :
> Yes, first I want to make a pairwise alignment to find the right 
> orientation of all sequences and after it's I'll align everything in a 
> multiple alignment.
>
> I meant this message:
> http://lists.open-bio.org/pipermail/biopython/2005-May/002656.html
>
> So you're right they talk about the printed output.
> I try to get the score.
> Thanks for the link!
> Stefanie
>
> ----- Original Message ----- From: "Peter" 
> <biopython at maubp.freeserve.co.uk>
> To: <lueck at ipk-gatersleben.de>
> Cc: <biopython at lists.open-bio.org>
> Sent: Sunday, November 23, 2008 2:32 PM
> Subject: Re: [BioPython] ClustalW Multiple Alignment
>
>
>> On Sun, Nov 23, 2008 at 1:21 PM, <lueck at ipk-gatersleben.de> wrote:
>>> Hi!
>>>
>>> I want to align several sequences under the allowance to check 
>>> whether there
>>> are reversed complement.
>>>
>>> That means I align all sequences against a reference sequence and then
>>> reversed complement them and align them again. After that I want to 
>>> compare
>>> the score and choose the better once.
>>
>> If I have understood you correctly, you have one reference nucleotide
>> sequence, and many nucleotide sequences of unknown orientation. For
>> each sequence you want to do a pairwise alignment against the
>> reference to decide which orientation matches best (forwards, or
>> reverse complement).
>>
>> So you want to lots of pairwise alignments. Perhaps ClustalW is not
>> the best choice - maybe use EMBOSS needle? You could also try
>> Biopython's Bio.pairwise2 module.
>>
>>> Now my question:
>>>
>>> How can I get the score?
>>
>> What score exactly are you looking for?
>>
>>> Unfortunately it's not in the dnd file.
>>
>> The dnd file from clustalw is just a tree, there is no score.
>>
>>> In an old message of this mailing list, it's was written that it's
>>> in the log file. Does this has been removed?
>>
>> What log file? I didn't think clustalw wrote a log file. It could be
>> in the standard output printed to screen...
>>
>> What old message on the mailing list are you refering to? Could you
>> link to it in the archive maybe?
>> http://lists.open-bio.org/pipermail/biopython/
>>
>> Peter
>>
>
> _______________________________________________
> BioPython mailing list - BioPython at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biopython