Train / Fruit fly multisensory circuit (FlyWire)

Odour T-maze with a delay

Can the circuit link an odour to a turn, and can it still do so when the odour is gone before the decision?

Loading 3D view…

Fly ·

Real wiringReadout from generation 0

Learning curve

press Train

No generations yet. Press Train: every generation adds a point for the best candidate, the population mean and the held out test.

Untrained (generation 1 mean)–

Best on training–

Held out, current readout–

Inside the circuit

live, current readout

readout neuronsfiring rateweight

All spikes, last 2 seconds

Saved runs

tick runs to overlay their held out curves

No saved runs yet. Train the real circuit and a control, save both, and compare their curves here.

What is being trained

Each trial starts with one of two odours: A (ORN DM1, fruity esters) or B (ORN DA2, geosmin). After a delay the fly reaches the junction, and the LC16 neurons signal that it is time to choose. Odour A means left, odour B right.

Training uses no delay and a 100 ms delay. The held out test uses 50 and 300 ms. The model has leaky neurons, no learning and no built in persistent activity, so a long delay is expected to erase the cue.

The benchmark found a more basic limit first: in this circuit the two odours barely reach the 21 output neurons (olfactory receptor neurons are at least three synapses away from them, and few strong paths survive the cut), so no readout can tell A from B, even with no delay. Every circuit stays at chance. A larger circuit with the lateral horn and mushroom body output neurons is the next step.

Inspired by: Odour choice assays in T-mazes (Tully and Quinn 1985).

How to read the result

Fixed: the wiring, the synapse counts, the transmitter signs and the neuron model (Shiu et al. 2024). Learned: 22 readout numbers.

A rising curve alone proves little: a trained readout can often make use of any network. The evidence is in the comparison. Train the real circuit, save it, train a rewired control, and overlay the two held out curves. If the real wiring learns faster or generalises better, the specific map carries information the controls lack.

Senses: Odour A → ORN DM1; Odour B → ORN DA2; Junction reached → LC16 (go signal). Readout neurons: DNa01 left, DNa01 right, DNa02 left, DNa02 right, DNa03 left, DNa03 right, DNb05 left, DNb05 right, DNa11 left, DNa11 right, DNp01 left, DNp01 right, MDN left, MDN right, DNp09 left, DNp09 right, DNg13 left, DNg13 right, DNp03 left, DNp03 right, MN9 (proboscis).

Training uses a 0.25 ms simulation step (the lab uses 0.1 ms); firing rates in these circuits differ by about 2 percent. Dataset: Subcircuit of FlyWire v783: inputs HSE, HSN, HSS, LPLC1, LC4, LPLC2, LC16, ORN_DM1, ORN_DA2, sugar/water, bitter; outputs DNa01, DNa02, DNa03, DNb05, DNa11, DNp01, MDN, DNp09, DNg13, DNp03, CB0701; paths with at least 40 synapses per step. About this dataset