Support for large genomes

[Tsaranoga]

Hello,

is it planned to support genomes that are larger than 2.1GB? (seeing #468)
at first glance intervaltree to int64_t and using pickle instead of msgpack would be a solution?
If yes i could also help with that, but getting alone through the code seems much work.

Cheers,
Tobias

[Tsaranoga]

made a pull request

[gemygk]

Thank you @Tsaranoga for your effort on PR #477. I have initiated git workflow builds, and they were successful.

Could you please provide a test dataset so I can test this internally? If it's publicly available, that would help.

Thanks,
Gemy

[Tsaranoga]

I can check for a public available one, the one i used is still unpublished.
for the int64 thing you can just use the human genome essentially and the second bug use any genome with an annotation from EviAnn (https://github.com/alekseyzimin/EviAnn_release)

[swarbred]

@Tsaranoga there is no issue running mikado on genomes over 2.1GB, we run regularly on wheat (16GB). The #468 query was in relation to chromosome size i.e. a single sequence over >2.1G mikado will work fine on the human genome