Please support CuffLinks GTF - this is a pretty important format and would help us get more users if we support it.
I attach an example.
Note that some "extra feature" fields are quantitative. It would be good to ensure that these are read in as floats not character strings and that they be available for the "color by score" feature. Note also that individual exons have scores and that the transcript as a whole has a score.
The attached file (Cufflinks_6K.gtf.gz) and its tabix index is from blueberry genome:
Vaccinium corymbosum genome versions V_corymbosum_scaffold_May_2013
Access it via private Loraine Lab RNASeq QL server http://transvar.org/illumina/ using your transvar user name and password.
I also attach an image showing how the file looks when parsed into IGB.
Note that CuffLinks probably has a non-standard implementation of GTF. However, the CuffLinks software is very, very widely used and even if the format is non-standard, IGB needs to support it.
Note that the second field is "Cufflinks" which is one way IGB can tell that the file is Cufflinks GTF.
Another option would be for the parser to detect that there are "exon" features that are not part of any parent "sim"
When that happens, IGB could open a window that asks the user about the format what the grouping feature can be.