microbiomedata / context-collaboration

Resources for data-driven and collaboratively refined mappings between ontology classes like ENVO:00002259 ‘agricultural soil’ to MIxS environmental context questions like env_medium, subsetted according to MIxS environmental extensions like Agriculture.

context-collaboration

Resources for data-driven and collaboratively refined mappings between ontology classes like ENVO:00002259 ‘agricultural soil’ to MIxS environmental context questions like env_medium, subsetted according to MIxS environmental extensions like Agriculture.

Website

https://microbiomedata.github.io/context-collaboration

Repository Structure

• examples/ - example data
• project/ - project files (do not edit these)
• src/ - source files (edit these)
  • context_collaboration
    • schema -- LinkML schema (edit this)
    • datamodel -- generated Python datamodel
• tests/ - Python tests

Developer Documentation

Use the `make` command to generate project artefacts:

• make all: make everything
• make deploy: deploys site

Credits

This project was made with linkml-project-cookiecutter.

context-collaboration's Issues

project asserted relations from pivoted ubergraph query onto subclasses?

opus gives verbose background knowledge comments on all mappings

are the current MIxS environmental extensions even well suited for NMDC?

How evenly are they distributed within NCBI Biosamples?

Do we have some other logical breakdown of types of Biosamples? Are the extensions good for anything other than requiring and recommending groups of slots?

extend llm emitted mappings over their subclasses

automatically check agreement between class slugs and labels in llm output

add categorical prioritization and stratus tracking back into this schema

the merged annotations/relations SAPRQL output may be close to opus' context limit

create tools for pivoting a LongTable to a wide voting table. combine agent and timestamps into one column.

combination columns:

• agent; timestamp
• confidence; comment

find patterns that can be accounted for with SAPRQL or oaklib queries

It would be nice to present a set of mappings to some tool and have it say "I'm surprised you didn't include XYZ. It fits into this pattern I noticed"

Or the tools could show some visualization that projects the mappings onto a a scoped background of the ontology

or the tools could just generate the queries themselves