Genomic Coordinate System

[ahwagner]

We have not indicated recommendations for how genomic locations are stored in genomic databases, and we have similarly not provided a recommended exchange format for gene fusion data between systems.

I think that these recommendations are out-of-scope for the first phase of this work. Our initial objective is only to define the information that must be captured to represent a gene fusion, and the choice of representation in a resource is a concern only for the implementers, so long as the representation is documented and applied consistently.

My personal recommendation for the choice of coordinate system is inter-residue, as used in the GA4GH VRS and BED standards and the NCBI SPDI standard. This is a useful system both for computation as well as unambiguous interpretation of locations (not interpreted differently for different variant types such as insertion/deletion/substitution). For more on this subject, see Figure 3A of the VRS manuscript and associated text.

This has the added benefit of clear visual representation of junction boundaries, as seen in the UCSC Genome Browser:

Recommendations and arguments for the use of a particular coordinate system are welcome here, and will be used to help define a future exchange format.