Re: GFF3 Reader

Craig

I think I fixed this problem but not sure if it is in the current
released jar files. The < logic was wrong where 0 wasn't be treated
properly. I will send you an update jar to see if the problem goes
away.

Scooter

On Fri, Feb 11, 2011 at 5:54 AM, Craig Berry
<[email protected]> wrote:
> Can I just have someone validate this logic in the GFF3Reader for me
> to see if this is a bug or not.
>
> If I have a GFF3 file with the following line:
>
> chrI    SGD     repeat_region   1       62      .       -       .       ID=TEL01L-TR;Name=TEL01L-TR;Note=Terminal%20stretch%20of%20telomeric%20repeats%20on%20the%20left%20arm%20of%20Chromosome%20I;dbxref=SGD:S000028864
>
> When parsing the file then, the class calls Location.fromBio using the
> start 1, end 62 and strand –ve.
> Since the strand is –ve it needs to convert the positions to negative
> values and reverse the start and end. However, as the javadocs
> explains:
>
> “In biocoordinates, the start index of a range is represented in
> origin 1 (ie the very first index is 1, not 0),  and end= start +
> length - 1.”
>
> So before the end is reassigned its value is reduced by 1 and then
> negated: e = - ( start – 1) With a start value of 1 as in this case,
> the end then becomes 0, such that the range now runs -62 to 0.
>
> This causes a problem when adding this Feature to the feature
> collection since it considers a position of value 0 to be on the +ve
> strand, such that when Location.plus() is called the check for a
> negative location (i.e. both start and end being < 0) returns false
> and so you end up trying to create a Location with a –ne start
> position but a +ve end, which throws an IllegalArgumentException.
>
> So is there something fishy here or not? I’m assuming that the GFF
> content is valid.
>
> Thanks in advance
>
> Craig
>
> _______________________________________________
> Biojava-l mailing list  -  [email protected]
> http://lists.open-bio.org/mailman/listinfo/biojava-l
>
>

_______________________________________________
Biojava-l mailing list  -  [email protected]
http://lists.open-bio.org/mailman/listinfo/biojava-l