[Bioperl-l] Bio::Restriction::Analysis feature merge

[Peter Blaiklock]

On Tuesday 09 December 2003 18:39, Rob Edwards wrote:
> Peter,
>
> I have a few comments on these revised modules:
>
> 1. Does circularity work OK? It fails the tests I devised at the moment
> (there is an MwoI site near the 0 point on t/data/dna1.fa)

   It does now. 
http://www.restrictionmapper.org/bioperl/Analysis.pm

> 2. It doesn't seem to handle multiple digests - did you add this in and
> am I missing it?

   Sorry I wasn't clearer before. Analysis->get_digest can be passed an enzyme 
name, an Enzyme object or an EnzymeCollection object. It returns an array of 
hashes with each hash corresponding to a restriction fragment. The hash keys 
are the same as from the 'fragment_maps' method, i.e. 'start', 'end' and 
'seq'.

> 3. Do you have other test scripts that you have run on these so I can
> compare?

   http://www.restrictionmapper.org/bioperl/ratest.pl
 
> 4. I don't like the idea of reverse complementing the target sequence.
> If you are using a large target sequence (e.g. 1 Mbp or more) this will
> be horrible. It would be a lot better to reverse the enzyme and should
> be the same (of course, you'd have to reverse the complementary strand
> of the non-palindroimic enzyme).

   It's easier to debug this way and I suspect that the large number of Enzyme 
objects is our worst performance bottleneck anyway. But if you want to 
reverse complement the recognition site instead I will write tests for the 
different possibilities (cut before site, cut in site, cut after site).

> 5. The newer version of Enzyme.pm (in cvs) doesn't require Storable.pm,
> and this is better than demanding it be there.

   Go ahead and get rid of it.

>
> If you are done working on this I'll fix it up and alter the method
> that gets the cut site to move that into Enzyme.pm so that Analysis
> just works with the integers, and add the changed modules into cvs
> before the weekend. If you have made any other changes can you email
> them to me so I can incorporate them too?
>
> Thanks
>
> Rob
>
> On Sunday, December 7, 2003, at 08:32  PM, Peter Blaiklock wrote:
> > Hi
> >
> >    I modified my new Bio::Restriction::Analysis module to incorporate
> > Rob's
> > changes, including the use of index to locate nonambiguous sites and a
> > digestion method that can handle multiple enzymes.  Please let me know
> > about
> > any problems, suggestions, bugs, etc. I should have tests ready in a
> > day or
> > two.
> >
> >  http://www.restrictionmapper.org/bioperl/Analysis.pm
> >  http://www.restrictionmapper.org/bioperl/Enzyme.pm
> >  http://www.restrictionmapper.org/bioperl/base.pm
> >
> >  Peter Blaiklock
> >

Peter Blaiklock