Overview

Virtual Fly Brain is a knowledge base for the Drosophila melanogaster nervous system. It brings together data produced by many different laboratories — electron microscopy reconstructions and the connectomes derived from them, confocal images of driver line expression, single-neuron and clonal morphologies, and single-cell transcriptomics — and makes them queryable together rather than one study at a time.

Two things make that work. Images are registered onto a small set of standard templates, so material from different studies sits in the same coordinate space and can be compared directly. And everything is classified with terms from the Drosophila Anatomy Ontology, so a query for a cell type returns data from every study that reported it, whatever that study happened to call it.

What you can ask

• Find a cell type or a region, and see every neuron, image and dataset VFB holds for it.
• Follow connectivity — what synapses onto a neuron, what it synapses onto, and how that differs between connectomes.
• Go from a neuron to a reagent. Given a cell type, find the split-GAL4 or GAL4 lines reported to label it, and the stock that carries them.
• Search by shape. NBLAST finds morphologically similar neurons across datasets, including matching an EM reconstruction against light-microscopy expression patterns.
• Compare across datasets, because the same ontology terms are applied to all of them.

Where to go next

• Concepts — the ideas the data model rests on: cell types, templates, bridging registrations, similarity scores and confidence values.
• Data — what VFB actually holds, dataset by dataset, and where each came from.
• Website features — how to drive the interface: search, Term Info and the viewers.
• APIs and Tutorials — the same data from Python, R or an AI assistant.

A tour of the site

The walkthrough above covers the main interface. Note that it predates some later changes to the site, so a few details will differ from what you see now.