[Biojava-l] [Biojava-dev] A question about multiple alignment
Andreas Prlic
andreas at sdsc.edu
Sun Sep 18 23:50:27 UTC 2011
Hi Shahab,
Sounds like you want to use an identity matrix for the alignment..
Andreas
On Sat, Sep 17, 2011 at 3:28 PM, Shahab Kamali <skamali at cs.uwaterloo.ca> wrote:
> Thanks Andreas,
> I want two components that have different names to have 0 alignment score.
> My application is not about bio-compounds,so I can use anything else rather
> than ProteinSequence and AminoAcidCompound. I just need to align sequences
> of arbitrary alphabets. Could you suggest me a solution please?
> Thanks a lot,
> Shahab
>
> Quoting Andreas Prlic <andreas at sdsc.edu>:
>
>> Hi Shahab,
>>
>> did you take a look at the substitution matrix, if it is scoring your
>> sequences according to your expectation? Looks like in your
>> theoretical example the alignment of B and D is favorable, i.e. it has
>> a positive alignment score..
>>
>> Andreas
>>
>>
>> On Fri, Sep 16, 2011 at 10:56 AM, Shahab Kamali <skamali at cs.uwaterloo.ca>
>> wrote:
>>>
>>> Hi,
>>> I am using BioJava in a pattern mining project. I want to align a set of
>>> relatively short sequences. For example to align {"ABCE", "ABCE", "ADE",
>>> "ADE").
>>>
>>> This is a part of my code:
>>>
>>> SubstitutionMatrix<AminoAcidCompound> matrix = new
>>> SimpleSubstitutionMatrix<AminoAcidCompound>();
>>> GuideTree<ProteinSequence, AminoAcidCompound> gt = new
>>> GuideTree<ProteinSequence,
>>> AminoAcidCompound>(lst,Alignments.getAllPairsScorers(lst,
>>> Alignments.PairwiseSequenceScorerType.GLOBAL, new
>>> SimpleGapPenalty((short)0,(short)0), matrix));
>>> Profile<ProteinSequence, AminoAcidCompound> profile =
>>>
>>> Alignments.getProgressiveAlignment(gt,Alignments.ProfileProfileAlignerType.GLOBAL,
>>> new SimpleGapPenalty((short)0,(short)0),matrix);
>>>
>>> The result of the above code is:
>>> ABCE
>>> ABCE
>>> AD-E
>>> AD-E
>>>
>>> But what I need is
>>> A-BCE
>>> A-BCE
>>> AD--E
>>> AD--E
>>> or
>>> ABC-E
>>> ABC-E
>>> A--DE
>>> A--DE
>>>
>>> Do you have any suggestion?
>>> Thanks,
>>> Shahab
>>>
>>>
>>>
>>> _______________________________________________
>>> biojava-dev mailing list
>>> biojava-dev at lists.open-bio.org
>>> http://lists.open-bio.org/mailman/listinfo/biojava-dev
>>>
>>
>
>
>
>
>
More information about the Biojava-l
mailing list