Uberon BOT module is huge; recommendations to shrink?

[matentzn]

Embedding UBERON in species specific phenotype ontologies is exceedingly hard because even small signatures lead to enormous bottom modules. For example, the import of UBERON for hp (33 MB) is larger than hp itself (27 MB). This is a problem; it makes building and QC processes long and merged artefacts huge. Does anyone have an idea, ever so hacky, to deal with this? Obviously UBERON is highly connected with many relations connecting all sorts of branches; we cant change that, and we need those for querying, inference and all. Still feel something should be done.

[cmungall]

Just take part-of/develops-from

…

On Fri, Feb 1, 2019 at 2:20 PM Nico Matentzoglu ***@***.***> wrote: Embedding UBERON in species specific phenotype ontologies is exceedingly hard because even small signatures lead to enormous bottom modules. For example, the import of UBERON for hp (33 MB) is larger than hp itself (27 MB). This is a problem; it makes building and QC processes long and merged artefacts huge. Does anyone have an idea, ever so hacky, to deal with this? Obviously UBERON is highly connected with many relations connecting all sorts of branches; we cant change that, and we need those for querying, inference and all. Still feel something should be done. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1478>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADGOZ1TBHPqiRp8gqD288MDs78MFTF5ks5vJL2lgaJpZM4afLv8> .