#268 GFF support incomplete?
Hello, developers! It's me again ;-)
It seems to me that GFF format support right now is incomplete. The problem can be clearly seen at one simplest 2-segment Drosophila R5 track attached. It is GFF, so segments in this file can have scores, right? So, I suggest that I should have option to see it as segments of different height (appropriate to score of each segment etc). Instead of it I see the boxes of fixed height and all options concerning the graph changing are not active.
The problem with BED is quite the same. example.bed contains the same two segments with scores. IGB displays coordinates and IDs quite correctly but doesn't change the height of each segment depending from it's score.
Second (rather cosmetic problem). The font size of ID is autodetected, I cannot fix it to the desired size. Probably it is good, but it would be resonable to have all IDs in the frame written by the same font size.
Thank you in advance anyway, your software is cool enough!
Discussion
Hi Youri,
We currently have the option to "Color by Score" which you may find helpful. See https://wiki.transvar.org/display/igbman/Customize+annotation+colors+using+Color+By for more details. We have an experimental option to fix annotation label sizes, but we have not yet made this option available. Stay tuned at http://bioviz.org/igb/news.html and read the release notes to see when this option becomes available. I suspect we will roll it out sometime in the next two months, but cannot give you a date at the moment. Thank you for the suggestions, we will take them into consideration.
Best Regards,
David
Hi Youri,
I need to add some additional comments to ensure you are able to get your current files to be visualized using our "Color by Score" feature. The feature unfortunately does not work correctly for the GFF format (this will be fixed in our next release), but you can work around this issue by using the gff3 format as an alternative.
You could also use your bed file to access this feature; however, your current file specifies type=bedDetail in the trackline, and this is causing the score field to be incorrectly mapped. If you remove the trackline from this file you will have no issues. Technically, you specification of the bedDetail type was not a violation of the spec, and I will patch this mistake in our parser in our next release as well.
I have attached a screenshot of the feature working as expected with your bed file with only the trackline removed.
I apologize for the mistakes in our parsers related to this feature, and hope these workarounds will help you.
FYI: If you are converting from GFF to BED you should be aware BED uses zero-base (interbase) coordinates but GFF uses one-based coordinates.
Best Regards,
David