Smith-Waterman against BLAST - Need Biologists' help!

[EngineerInBiology]

Hi to everyone,
I'm new to the forum and new to sequence alingment as well. I'm working on a FPGA hardware accelerated custom solution for exact local alignment.

Since I'm not a biologist, I found some difficulties in finding a "real" examples of two biological sequences where BLAST is missing relevant homologies due to its heuristical limits. Better if this example is a "resounding" one and that would be otherwise found using SW.

Do you know some example? I would really appreciate your help, since you can imagine how difficult can be for a non-biologist to find such information.

Thanks in advace. If I'have not been clear or you need more information, please let me know.

Alessio

[krobison]

I'm a bit surprised this ground is being plowed yet again; there were several commercial products in this space about 15 years ago (TimeLogic, Compugen & Paracel), but as microprocessors kept ramping up these companies faded or mutated away from the market.

But, it sounds like a fun project. I don't know of good examples off-hand, but one way to find them would be to use SCOP as ground truth for homology, and then run a bunch of BLAST searches to identify cases in which BLAST can't find a plausible alignment between two members of the same SCOP family -- then run SW to identify which of those pairing SW can pick up.

You might also find this paper useful.

Good luck!

[EngineerInBiology]

Thanks a lot for your prompt reply!!!
I did a research in literature about SW-BLAST comparison and I found many articles, including the good one you mentioned (thanx).

I was wandering if there were some "famous" example or something one has found in his works, let's see if anyone from the forum have some additional information or advice.

Using SCOP is a good idea anyway, maybe more realizable than I thought when first reading those article (I think it will take a little time, at least for me).