any human differs less than 1 % (although it really depends on how you count differences). It would make sense to store one reference, and then everybody is st

any human differs less than 1 % (although it really depends on how you count differences). It would make sense to store one reference, and then everybody is stored as a delta relative to that (https://en.wikipedia.org/wiki/CRAM_(file_format) https://en.wikipedia.org/wiki/CRAM_(file_format) and https://en.wikipedia.org/wiki/Compression_of_genomic_sequencing_data https://en.wikipedia.org/wiki/Compression_of_genomic_sequenc...)