[ssml] Re: pdb-l: All-vs-All Global protein seq alignment score

[Karsten Suhre]

Hi Manisha,

  I am not sure what your problem is exactly, but computing pairwise global 
alignments of distantly related proteins makes little sense. In that case I 
would rather - for each pair - add a number (5-10) closer homologues for each 
protein (50-70% homology) and compute a multiple alignment for that set. Here 
T-Coffee would be my program of choice (it computes multiple alignment 
scores). On the other hand, if all of your 400 sequences are close 
homologues, I would use MUSCLE to align all 400 in one go. I would then use 
protdist from phylip to compute phylogenetic distances between all pairs if 
that is of use for your purpose.

  Good luck,

  Karsten.

On Monday 10 January 2005 20:47, Goel, Manisha wrote:
> Hi All,
>
> I have a database of about 400 squences, which I want to compare
> All-vs-All to get the global alignment score of each protein sequence
> pair.
> I tried using bl2seq but that gives me local alignments.
> What other program would suit this function best ?
> I would probably like to have the flexibility of being able to use
> various matrices.
>
> Thanks a lot,
> -Manisha