I'm wondering, from the point of view of an annotation library like Ensembl, does it make any sense for a locus to be inside a gene but not inside any transcript?
The definition of the term gene is quite broad:
Gene: http://www.sequenceontology.org/browser/current_svn/term/SO:0000704
A region (or regions) that includes all of the sequence elements necessary to encode a functional transcript. A gene may include regulatory regions, transcribed regions and/or other functional sequence regions.
Transcript: http://www.sequenceontology.org/browser/current_svn/term/SO:0000673
An RNA synthesized on a DNA or RNA template by an RNA polymerase.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
The inclusion of regulatory regions seems, I think, different from how Ensembl defines genes. It's also confusingly vague: would you include all eQTLs as part of a gene?
The key point there is the wording that says "may include" but does not have to. I found that different communities interpret the specification quite differently. For example annotations for bacterial genomes will follow different standards than those for mouse etc