[Bioperl-l] Help with Bio::SeqIO

[Chris Fields]

On Nov 5, 2007, at 10:54 AM, Jay Hannah wrote:

> On Nov 5, 2007, at 11:03 AM, Hilmar Lapp wrote:
>> I agree that there should be a meaningful default that results in
>> "doing the right thing" in most cases if the user doesn't intervene.
>> I'm not sure I understand all the details, but it sounds sorting or
>> not sorting should depend on the split location type unless the user
>> overrides it by argument. That's what you're suggesting, right?
>
> If someone knows why spliced_seq() should ever sort then I'm
> suggesting we add a test demonstrating a useful example of that.
>
> If no one has a useful example of when you would want spliced_seq()
> to sort then I'm suggesting we remove the sorting altogether and
> nosort goes away.
>
> I can provide/add many examples where sorting is bad. I do not know
> of a case where sorting is good.
>
> j
> http://www.bioperl.org/wiki/User:Jhannah

The behavior would be based on the current use of 'join', 'order',  
and 'bond' (the latter in GenPept records).  I documented some cases  
here a while back:

http://www.bioperl.org/wiki/BioPerl_Locations#Split

chris