Virtual Fly Brain

University of Edinburgh School of Informatics
United Kingdom of Great Britain and Northern Ireland

About

Launched: 2010
Record Updated: Sep 30, 2026
Discovery system
Open scholarly dataset
Virtual Fly Brain (VFB) is a FAIR, queryable knowledge base and 3D atlas of the Drosophila melanogaster nervous system. It registers tens of thousands of neuron, clone and gene-expression images onto standard template brains, classifies them with the Drosophila Anatomy Ontology, and integrates connectomic, transcriptomic and reagent-expression data from across the field. Researchers can search by name, anatomical location, connectivity or image similarity (NBLAST), browse 3D reconstructions and EM connectomes, and query the underlying knowledge graph via Python (VFB_connect), REST/Cypher APIs, or an LLM chat interface. VFB is also the long-term public host for several community connectomics datasets (e.g. FAFB) via CATMAID, preserving access when original hosts go offline.

Domain Relevance

  • Biological sciences

Mission

To give Drosophila neuroscientists a single, queryable, FAIR view of the fly nervous system, unifying neuroanatomy, connectomics and gene-expression data that would otherwise sit in disconnected image collections, ontologies and papers, so that anatomical and circuit-level questions can be answered by search and query rather than manual literature review.

Key Achievements

Integrated eight major connectomics datasets (hemibrain, MANC, Optic Lobe, FlyWire, BANC, MaleCNS, and two larval EM datasets) into a unified, queryable connectome resource. Shipped VFBchat, an LLM-based natural-language query interface over the VFB knowledge base, and a companion Model Context Protocol (MCP) tool for direct use from other AI assistants. Continued as the long-term public host for community EM connectome datasets via CATMAID, including preserving access to FAFB after its original host retired the service.

Technical Attributes

Maintenance Status

Actively Maintained

Open Code Repository

Implemented

Technical Documentation

Implemented

Code License

Implemented

Open API

Implemented

Open Data Statement

In Progress

Open Product Roadmap

In Progress

Technical Attribute Statements

Programming Languages

  • java
  • javascript
  • python
  • typescript

Technology Readiness Level

  • Actual system proven in operational environment

Code Licenses Used

  • Apache License, Version 2.0
  • GNU General Public License (GPL)
  • MIT License

Content Licensing

  • Creative commons licenses

Standards

Metadata

  • JSON
  • RDF
  • Web Ontology Language (OWL)

Community Engagement

Code of Conduct

In Progress

Community Engagement

In Progress

Community Statements

User Contribution Pathways

  • Contribute to code
  • Contribute to documentation
  • Contribute to education or training
  • Contribute to user research or user testing

Engagement with Values Frameworks

  • FAIR Guiding Principles for Scientific Data Management and Stewardship

Additional Information

Organizational History

Virtual Fly Brain began in 2010 through BBSRC-funded work at the University of Edinburgh to build a queryable atlas of Drosophila neuroanatomy, growing out of earlier ontology and image-registration research (the Drosophila Anatomy Ontology and FlyBrain image atlases). It has since expanded through continued BBSRC and Wellcome Trust funding into a multi-site collaboration spanning the University of Edinburgh, University of Cambridge, and the MRC Laboratory of Molecular Biology, integrating an increasing number of connectomics datasets (hemibrain, MANC, Optic Lobe, FlyWire, BANC, MaleCNS and larval EM data) alongside its original neuron and gene-expression image collections, and now also serves as a long-term public host for community EM connectome datasets such as FAFB via CATMAID.

Organizational Structure

Business or Ownership Model

Distributed institutional partnership

Funding

Primary Funding Source

  • Contributions

Revenue Sources

  • Grants (private)
  • Grants (public)
  • In-kind contributions

Financial Reporting Level

  • Provider

Fiscal Host

University of Edinburgh