isoform_id generation utilize some form of binary encoder based on exons at preprocessing.py
This can often results in novelIsoform_ naming that is repeated across multiple genes and chr. Non-unique transcript IDs can cause issues with annotation parsers.
Recommend to append gene name, gene ID, or generate a unique hash based on genomic location to novel isoforms.
Thank you for the hard work!
isoform_id generation utilize some form of binary encoder based on exons at preprocessing.py
This can often results in novelIsoform_ naming that is repeated across multiple genes and chr. Non-unique transcript IDs can cause issues with annotation parsers.
Recommend to append gene name, gene ID, or generate a unique hash based on genomic location to novel isoforms.
Thank you for the hard work!