[Bioperl-l] Bio::Align::DNAStatistics ka/ks calculation bugs?

[Jason Stajich]

That's right.   see the PAML HOWTO - $dna_aln can be slices of an  
alignment.

# set the alignment object
$kaks_factory->alignment($dna_aln);

# run the KaKs analysis
my ($rc,$parser) = $kaks_factory->run();
my $result = $parser->next_result;
my $MLmatrix = $result->get_MLmatrix();

On Nov 1, 2005, at 7:00 PM, Bingshan Li wrote:

> Thanks, Jason, for this reply and the other for Ganbank parser. For  
> Yn00, can we pass a slice of alignment to it so that I can  
> calculate ka/ks for sliding windows? I ran PAML as a stand alone  
> program but never used it in perl script.
>
> Thanks a lot!
>
> -bs
>
> On Nov 1, 2005, at 4:45 PM, Jason Stajich wrote:
>
>
>> I suggest Codeml and YN00 - I assume you are only doing this for  
>> pairwise comparisons and not along a tree?
>>
>> The dnastats  module calculates Nei-Gojobori so it will be  
>> different, but shouldn't be that different so something must be  
>> wrong with your code or the module's method.  Maybe you can post a  
>> bug to bugzilla.open-bio.org with a clear example of where it is  
>> going wrong and we'll take a look.
>>
>> I would just use YN00 for most pairwise - you can parse it with  
>> Bio::Tools::Phylo::PAML and automate running it with  
>> Bio::Tools::Run::Phylo::PAML::Yn00
>>
>>
>> -jason
>> On Oct 31, 2005, at 5:54 PM, Bingshan Li wrote:
>>
>>
>>
>>> Hi all,
>>>
>>> I used Bio::Align::DNAStatistics module to calculate ka/ks ratio  
>>> but I found the result is very different from the output by yn00  
>>> in PAML 3.14. I applied "calc_KaKs_pair" function on two  
>>> sequences to test it. The difference is pretty big (for one case  
>>> it gave ka/ks=1.35 while yn00 gave 0.37). Does anybody have same  
>>> experience? Is there any other methods I can use to calculate ka/ 
>>> ks ratios? I will also use it to calculate ka/ks in sliding windows.
>>>
>>> Thanks a lot!
>>> _______________________________________________
>>> Bioperl-l mailing list
>>> Bioperl-l at portal.open-bio.org
>>> http://portal.open-bio.org/mailman/listinfo/bioperl-l
>>>
>>>
>>>
>>
>> --
>> Jason Stajich
>> Duke University
>> http://www.duke.edu/~jes12
>>
>>
>>
>>
>
>

--
Jason Stajich
Duke University
http://www.duke.edu/~jes12