Some of the queries given to us at NeuroFly 2026
Real queries put to VFB at the NeuroFly 2026 workshop table, answered live and kept here as worked examples
At the VFB table during NeuroFly 2026 (Cologne, September 2026), people put real queries to us about their own neurons of interest. Each one below was answered live with the VFB MCP — the same tool the No-Code track teaches — and the answer, including any figure it produced, is kept here as a worked example.
These aren’t scripted demos: they’re genuine queries, the tool calls that answered them, and an honest account of what VFB’s data does and doesn’t say.
Index
| Query | Asked | Summary |
|---|---|---|
| Ring-shaped neurons in the fly connectome | 7 Sept 2026 | Which VFB neuron types have a ring-shaped morphology, and what does VFB hold on them |
| What feeds the moonwalker descending neuron? | 7 Sept 2026 | MDN’s strongest presynaptic partners, cross-checked across two independent connectomes |
| Split-GAL4 lines targeting PPL1 | 8 Sept 2026 | Why the whole-class query undercounts, and where PPL1’s real split coverage lives, subtype by subtype |
| Connections to Tm3 | 8 Sept 2026 | Tm3’s strongest recorded inputs, and why the whole-class connectivity query surfaces ontology umbrella terms first |
| Connectivity between L3 and Tm9 | 9 Sept 2026 | The per-pair synapse-count distribution of a canonical optic-lobe connection, and how consistently it holds across three connectomes |
| What neuromodulator data does VFB hold? | 9 Sept 2026 | Curated class assertions versus per-neuron EM transmitter predictions, and where the gaps are |
| MBONs, the ‘dopaminergic’ DNp32, and flight | 9 Sept 2026 | Weak MBON20/MBON35 input, three connectomes predicting three different monoamines, and output that goes to leg and jump motor neurons rather than wing or haltere |
Each entry records the query as it was asked, the VFB MCP tools and queries used to answer it, what came back, and the caveats — not just the headline numbers.