Currently we only support Ensembl. Maybe we could also test refseq and gencode.
This would required conversion for the chromosome name notation e.g and some other fields towards the end.
Check results for genes that are the same across different annotations
Check if the extra features (3utrs for example) are considered by alfa or this causes an issue.
Which rules do we expect to produce different results?
Make sure in the documentation that we mainly support Ensembl.
Mention we haven't tested extensively for all organisms so things might not work properly for rare organisms.
I think supporting gencode might be easier, not sure if we want to fight for RefSeq, though.
If you ask me - if we put out a workflow that supports ENSEMBL and gencode thats good enough.