[Biopython] Cross-Links in circular GenomeDiagram?

[Peter Cock]

On Thu, Nov 21, 2013 at 8:53 AM, Leighton Pritchard
<Leighton.Pritchard at hutton.ac.uk> wrote:
> Hi Ilya,
>
> On 20 Nov 2013, at Wednesday, November 20, 21:06, Ilya Flyamer wrote:
>>
>> By the way, another thing. Crosslinks between tracks in circular diagrams
>> also work in a weird way. You can see a picture here: http://itmag.es/2B8MM .
>> Why does it connect closely located regions with such huge crosslinks,
>> which go around the whole track? Why not connect them with arc going
>> counterclockwise (inside --> outside)?
>
> Peter wrote the crosslinks, but I think that this behaviour occurs because
> the motivation for including them was to represent connections on linear
> diagrams. On linear diagrams, it doesn't make sense to cross the origin
> (i.e. to go off the page to the left, then come back in on the right). The
> circular representation is currently, I think, a reapplication of the same
> logic in the circular context, rather than a rewrite specific to circular images.

Yes, that is a fair description of the current behaviour. This is something
I was wondering about working on, at least the the case where the
circular track is drawn as a full circle (not as a large arc with a pie
slice missing).

>>
>> And also crosslinks are hard to see under track features, but that
>> might be caused by the first issue.
>
> I'm not sure what you mean - do you mean that the angle at which
> the crosslinks come in can be so shallow that you can't separate
> them by eye?

Yes, extremely shallow links are hard to see, but there isn't much
we can do about that, is there?

Peter