2024-01-12

This is PMA's DBGI daily open-notebook.

Today is 2024.01.12

Todo today

Have a look at the DBGI discussion forum

- https://github.com/orgs/digital-botanical-gardens-initiative/discussions

Doing

Anticipated LOTUS

edges with very different weight
no true negatives
no edge weight for the negatives

They are models which are positive

Inductive models means that we cannot use node embedding per se

Find common features across biological sources and molecules. The wikidata page of these nodes could be a thing.

SciBERT

Creating a Anticipated Lotus page in the Dendron DBGI vault.

Textual description at least 10 words of both species and molecules.

This textual description needs also to be as complete as the existing ones. So a WD description would be OK if we keep name, INCHI, SMILES etc, but not the pubchem id for ex.

EMIKG Portal

Mono repo approach

e.g. google is a mono repo

https://github.com/enpkg/enpkg_full

https://en.wikipedia.org/wiki/Monorepo

Rust monoliths backend : handles the DB, the authorizations, Sample management, etc. No particular reason to do it in Rust except for the security reasons. Should be as integrated as possible.

Rust part listen to API calls from Python modules.

When you expect to have full control of a thing, e.g. you are the provider.

Created emikg-portal

A prefixed tree https://docs.rs/trie-rs/latest/trie_rs/ To integrate.