[Biopython-dev] Fwd: [Biopython] Exonerate Parser Error

[Zheng Ruan]

Forget to cc'd to dev list.

Hi Lenna,

I'm not quite sure about CoordinateMapper, but it seems to deal with
sequence files
with rich annotation like genbank. However, In our case, we are typically
not sure
about the coordinate correspondence between dna and protein sequence.
That's why
exonerate can help.

Thanks!


On Fri, Mar 21, 2014 at 4:54 PM, Lenna Peterson <arklenna at gmail.com> wrote:

> On Fri, Mar 21, 2014 at 3:32 PM, Zheng Ruan <zruan1991 at gmail.com> wrote:
>
>> Hi Bow,
>>
>> I have the same problem when trying to model codon alignment with
>> frameshift being considered. Basically, I have a CodonSeq object to store
>> a
>> coding sequence. The only difference between CodonSeq and Seq object is
>> that CodonSeq has an attribute -- `rf_table` (reading frame table). It's
>> actually a list of positions each codon starts with, so that translate()
>> method will go through the list to translate codon into amino acid. In
>> this
>> case, it is easy to store a coding sequence with frameshift events. And
>> it's not necessary to split the protein to dna alignment into multiple
>> part
>> when frameshift occurs. However, the problem now becomes how to obtain
>> such
>> information (`rf_table`). I find exonerate is quite capable of handling
>> this task, especially with introns in the dna. I do think an object to
>> store protein to dna alignment is necessary in this scenario.
>>
>
> Is the (still unmerged) CoordinateMapper the solution to this?
> http://biopython.org/wiki/Coordinate_mapping
> If so, let me know and I'll rebase and refresh the pull request.
> If not, I misunderstood the problem.
>
> Cheers,
>
> Lenna
>