A multi-agent question-answering tool that combines structured pathway knowledge from user-selected MINERVA projects with live web research (Perplexity) and synthesizes evidence using an LLM. It targets specialized questions in lipid metabolism and related tox/biology by grounding answers in curated MINERVA maps and corroborating sources.
Service Introduction: LLMapperino README
Workflow: 3-Agent Workflow (MINERVA ↔ Perplexity ↔ Synthesis)
Demo: Hosted pilot
Sub-stage: prototype
Development Cloud: http://81.169.225.178:5016/
Login Required: Email i.djidrovski@uu.nl for an app API key or provide your own OpenAI and Perplexity API keys in the UI.
TRL: TRL4
Type: pilot
Contact: Ivo Djidrovski (i.djidrovski@uu.nl)
API Type: Not available
Categories: To be added
Targeted Users: To be added
Relevant VHP4Safety Use Case: To be added
Provider: Utrecht University / VHP4Safety (pilot)
Citation: Not available
Version: 0.1.0
License: Apache-2.0
Source Code: https://github.com/senseibelbi/LLMapperino
Docker: https://github.com/senseibelbi/LLMapperino/blob/main/Dockerfile
Bio.tools: Not available
FAIRsharing: Not available
TeSS: Not available
RSD: Not available
Wikipedia: Not available