📢 biocypher / biochatter: Knowledge Graphs and Large Language Models
Knowledge Graphs (KGs) and Large Language Models (LLMs) are two highly synergistic and current topics in biomedical research. With our frameworks for both KG creation (https://biocypher.org) and LLM interaction (https://chat.biocypher.org), we aim to improve access to these workflows for the entire research community. Join our workshop to contribute to this timely topic!
🔬 What's in store:
- Integrate KG and LLM: Enhance user requests with KG data, cross-verify LLM outputs, and store results in KG for long-term use.
- Facilitate model chaining: Expand LangChain components and integrate APIs for gene expression and visual models.
- Establish benchmarks: Evaluate prompt sets for bioinformatics tasks and create LLM-orchestrated multi-stage workflows.
- Better UI integration: FastAPI support for smoother interactions.
- Tackle specific bioinformatics tasks: Cell type annotation with ontology-based recommendation systems and causal reasoning integration.
🧑🎓 What you should bring:
Both libraries are written in Python, and as such you should be familiar with Python programming and packaging. Further optional skills that could be useful for our topic include: Interest in and knowledge about modern generative AI models (prompt engineering, APIs), Knowledge Graph (or other database) technologies, TypeScript.
Project lead: Sebastian Lobentanzer,