This repository hosts structured XML data for Key Event Relationships (KERs) related to oligodendrocyte differentiation and myelination.
Note on Purpose: The primary output of this project is the machine-readable XML data generated directly from scientific literature using Large Language Models (LLMs). The XSLT visualization is provided solely as a tool to facilitate human validation of the AI's extraction accuracy.
This work is a contribution to the project: AI4AOP Managed by the European Commission's Joint Research Centre (JRC) and SCAHT, this initiative explores how artificial intelligence can accelerate the development and characterization of Adverse Outcome Pathways to support non-animal testing strategies.
The XML files in this repository are styled using XSLT to be human-readable. View the styled landing page here: https://ideaconsult.github.io/aopwiki_challenge/
The repository focuses on a specific AOP segment concerning developmental neurotoxicity:
-
KER 1977 → 2116: [Disruption of sodium channel gating kinetics] to [Decreased, oligodendrocyte differentiation].
- Weight of Evidence: Low-to-Moderate.
- Key Insight: Explores how altered neuronal firing patterns impact activity-dependent signaling required for OPC maturation.
-
KER 2116 → 2107: [Decreased, oligodendrocyte differentiation] to [Hypomyelination].
- Weight of Evidence: High.
- Key Insight: A direct mechanistic link where the failure of OPCs to mature results in an obligatory reduction of CNS myelin.
This project uses "Browser-Side Transformation" to display XML data:
- XML: Validated against the
aop-xmlnamespace (http://www.aopkb.org/aop-xml). - XSLT (
ker.xsl): A stylesheet that transforms raw XML into a clean, responsive HTML format with highlighted evidence levels and citations. - GitHub Pages: Used to host files as a web server, bypassing browser security restrictions for local XSL transformations.
- AOPMapper: Advanced mapping and visualization tool for Adverse Outcome Pathways.
- AOP Wiki: The central repository for AOP knowledge.
- AI4AOP: Boosting the Adverse Outcome Pathway Framework with AI.
Created for the AOP Wiki KER Challenge (2026).