caufieldjh
commented
1 year ago Hi @rickbeeloo, Glad to hear that OntoGPT may suit your needs! We don't currently have a way to pass an arbitrary set of ontologies directly on the command line, and that's partially because even two people using the same set of ontologies may have completely different use cases for them. Some small modifications to one of the existing schema will help, though. Here's an example:
id: http://w3id.org/ontogpt/rickbeeloo
name: rickbeeloo
title: rickbeelooTemplate
description: >-
A template for extracting ENVO, NCBITAXON, and FOODON
license: https://creativecommons.org/publicdomain/zero/1.0/
prefixes:
linkml: https://w3id.org/linkml/
rickbeloo: http://w3id.org/ontogpt/rickbeeloo
default_prefix: rickbeeloo
default_range: string
imports:
- linkml:types
- core
classes:
EntityContainingDocument:
tree_root: true
is_a: NamedEntity
attributes:
environments:
range: Environment
multivalued: true
description: >-
A semicolon-separated list of environmental terms.
taxa:
range: Taxon
multivalued: true
description: >-
A semicolon-separated list of taxonomic terms of living things.
foods:
range: Food
multivalued: true
description: >-
A semicolon-separated list of foods.
Environment:
is_a: NamedEntity
id_prefixes:
- ENVO
annotations:
annotators: sqlite:obo:envo
prompt: >-
the name of an environment.
Examples are lake, meadow, waterfall.
Taxon:
is_a: NamedEntity
id_prefixes:
- NCBITAXON
annotations:
annotators: sqlite:obo:ncbitaxon
prompt: >-
the name of a taxonomic name or species.
Examples are Bacillus subtilus, Bos taurus, blue whale.
Food:
is_a: NamedEntity
id_prefixes:
- FOODON
annotations:
annotators: sqlite:obo:foodon
prompt: >-
the name of a food, beverage, or ingredient.
Examples are ketchup, milkshake, grape jelly.
OK, so that's a full template, but feel free to adapt to your use case.
rickbeeloo
Hey @caufieldjh!
Awesome idea! This is probably exactly what we need (I hope :)). We have a bunch of free text fields that we want to standardize by mapping them to ontologies (ENVO, NCBITAXON, and FOODON). Reading the docs I suppose this should be possible, but I'm a bit lost in the docs.
Lets take this as example, in
test.txt:This is a sample from drinking water with E.coliWhich should give me something along the lines of:
I wonder what command should we run to do the above? I expected to be able to run something like
ontogpt --inputfile test.txt -t ENVO,NCBITAXON,FOODONbut looking here I should probably use one of the presets?Thanks!