Chris Kafer wrote:
>>mperr1 at hamon.swmed.edu (Michael Perry) wrote:
>> Thanks for the reply but I tried that first. What it appears to do is
> compare the entire sequence (as if it were a single ORF) in all 6
> frames. What I would like to do is have it blast the real ORFs >30
> amino acids individually. When I blast them manually, it brings back
> different hits than a blastx search of the entire sequence.
>> As an example, when I blast one of the 200 aa ORFs it returns an EST
> clone. When I blast the entire sequence, of which this ORF is just a
> part, it doesnt show up. Hopefully I explained that clearly!
>
You probably already realise this, but the fact that it comes
up in ones search and not the other is "meaningful" in the
BLAST view: what you have done is 'manually' trim one sequence
to a region, specialising in one particular region, and therefore
hits against that region now get a lower probability of being
random. [what I am saying is that this is what you expect].
[now for the plug]:
Although it does not do all of what you want, PairWise and
SearchWise can run a DNA query against a protein database allowing
for Frameshifts and to a certain extent, introns. The DNA2DNA
approach which you want we are developing but it is not released
yet. You can get PairWise and SearchWise from
http://www.sanger.ac.uk/~birney/wise/topwise.html
ewan
