[Bioperl-l] Why does Bio::DB::GFF::Feature::gff3_string swap start and stop coordinates??

[Chris Fields]

glimmer2/3 both assume the genome is circular by default (I'm  
assuming since Glimmer2/3 are used for bacterial genomes).  Acc. to  
the Glimmer3 release notes the detail file has the information in the  
header; from the Glimmer3 data used for tests:

Command:  /bio/sw/glimmer3/bin/glimmer3 -o 50 -g 110 -t 30 ../BCTDNA  
Glimmer3.icm Glimmer3

Sequence file = ../BCTDNA
ICM model file = Glimmer3.icm
Excluded regions file = none
List of orfs file = none
Truncated orfs = false
Circular genome = true
...

There are options available for glimmer3 (-L, -X) that specify a  
linear sequence or allow ORFs to extend past the end of the sequence  
analyzed (the latter assumes a linear sequence).

chris

On May 21, 2007, at 4:21 PM, Mark Johnson wrote:

>     That makes sense.  Is that behavior documented anywhere?  I'll
> feel like less of an idiot if it's not.  8)  Either way, if you're
> sure that's whats going on, I'll fix up the parser to handle that as a
> split location.
>
>> I think I know what it is.  If you mean these predictions:
>>
>> Glimmer2:
>>
>>     27    29263        6  [+1 L= 684 r=-1.187]
>>
>> Glimmer3:
>>
>> orf00001    29263        9  +1     9.60
>>
>> Glimmer2/3 are predicting a gene for a circular chromosome that
>> starts at 29263 and ending at +9 (+6 for Glimmer2, which leaves off
>> the stop codon).  Note in Glimmer2 detailed output the end is 29946
>> and the length of the sequence is 29940, so Glimmer2 artificially
>> extends the end of the sequence with part of the start.
>>
>> This is handled as a split location in bioperl and in most GenBank
>> files; the above would be a location string like 'join
>> (29263..29940,1..9)'.  If you switched the start and stop the
>> location would be '9..29263' which wouldn't be correct (and would be
>> a huge gene).
>>
>> chris
>>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioperl-l

Christopher Fields
Postdoctoral Researcher
Lab of Dr. Robert Switzer
Dept of Biochemistry
University of Illinois Urbana-Champaign