[gt-users] python tests and splitfasta

[Brent Pedersen]

On Thu, Apr 16, 2009 at 9:28 AM, Gordon Gremme <gremme at gmail.com> wrote:
> Hi Brent,
>
>> hi, i'm attaching a patch to update the python tests for recent change
>> to use create_new() rather than __init__()
>> i also added a property for attribs as i think that's convenient.
>
> thanks, but it leads to a test failure on my machine (Darwin i386):
>
> 488: gtpython: AnnotationSketch bindings (FeatureNode(Iterator)) : failed
>     [ problem: unexpected return code: 2 != 0
>       in: /Users/gordon/work/genometools/testsuite/stest_testsuite/test488 ]
>
> Any ideas?
>

that fails for me without the patch as well. i can't see what's going on.


>
>> i would like to add an option to the splitfasta command-line tool so
>> that the targetsize can be specified in basepairs rather
>> than MB, and another option to allow overlapping of sequences-- so it
>> can take a huge chromosome and e.g. split it into
>> 1megabase chunks with 10Kbp overlapping at the ends.
>>
>> i think the syntax would look like:
>>
>> gt splitfasta -splitbp -targetsize 1000000  -overlap 10000
>>
>> so splitbp is bool flag to tell it not to bp rather than mb.
>> does this seem reasonable and/or would a patch be accepted for adding that?
>
> Seems like a good idea to me! But I would rather not change the
> meaning of -targetsize with an additonal option. I think it would be
> better to use an additional option instead which excludes the usage of
> -targetsize (analog to -numfiles and -splitdesc). Something like that:
>
> gt splitfasta -splitbp 1000000 -overlap 10000
>
> Bonus points if -overlap works with -targetsize, too ;-)


ok. for now i've taken the easy way out and written a quick python
script to do this. i may have another look later.
-brent

>
> Gordon
> _______________________________________________
> gt-users mailing list
> gt-users at genometools.org
> http://genometools.org/mailman/listinfo/gt-users
>