INTERNATIONAL RESEARCH / PEER-REVIEWED ARTICLE

How scientists mapped a fruit-fly brain

Researchers mapped 139,255 neurons and 54.5 million synapses in one adult fruit fly at connection level. We explain what this map reveals and what it still cannot tell us.

ACADEMIC WORK REVIEWED

Neuronal wiring diagram of an adult brain

ResearchersSven Dorkenwald, Arie Matsliah, Amy R. Sterling and the FlyWire Consortium.

Nature · 3 October 2024 · 634:124–138 · Peer-reviewed research · Source language: English

Open the original publication Find on Google Scholar

ENGLISH REVIEW AND COMMENTARY: DIJITALPI

We explain the paper in plain English, discuss what it may mean for organisations and identify our own examples separately.

Read the explanation Go to 3 business examples

English review: 11 September 2026 · The source date appears in the citation above.

Published review

Source and editorial checks are complete. DijitalPi did not conduct a customer experiment for this paper.

LET'S READ THE RESEARCH TOGETHER

We first explain the researchers' question, method and findings. We then discuss how to interpret the results, clearly separating DijitalPi's commentary from the source.

01 / WHAT DID THE RESEARCHERS WANT TO UNDERSTAND?

What do we gain by mapping every connection in a fruit-fly brain?

The FlyWire Consortium brought the neurons and synapses of an adult Drosophila brain into one wiring map, creating a shared foundation for studying brain circuits.

Explaining the context · DijitalPi commentary

A connectome is a wiring diagram. It shows which neurons connect, but it does not by itself prove which connection produces a particular behaviour.

Although a fruit fly has a small brain, it supports research into vision, navigation, learning, memory and social behaviour.

02 / HOW WAS THE RESEARCH CONDUCTED?

How did they test the question?

Researchers examined a single adult female fruit fly with high-resolution electron-microscopy images. Computer vision and deep-learning-assisted methods separated neurons in the images, while people checked connections and errors.

Understanding the method · DijitalPi commentary

Electron microscopy provides synapse-level detail, but the raw images form a complex, unsegmented web. Automation speeds segmentation; community proofreading improves reliability.

The study is based on one adult female fly. Biological variation between brains, imaging errors and incomplete connections limit what can be generalized.

03 / RESEARCH FINDINGS

One wiring diagram brought together 139,255 neurons and 54.5 million synapses.

The researchers presented a connectome combining 139,255 neurons, about 54.5 million chemical synapses and information paths across brain regions. The data can be searched and visualised.

Neurons: 139,255Neurons mapped in one adult female fruit-fly brain.
Synapses: 54.5 millionApproximate number of chemical connections between neurons.

Source: Nature research article and DOI ↗

04 / DIJITALPI'S EXPLANATION

How should we interpret these findings?

The map lets scientists select individual circuits and test them with behaviour and neural-activity experiments.

AI acts as a scale-up tool for turning large image collections into workable data. Reliability still depends on human review and data records.

Seeing a network’s connections is not the same as seeing how it operates over time. Structural wiring, live activity and behaviour should not be presented as one claim.

05 / CONCLUSION AND OPEN QUESTIONS

What did we learn, and what do we still not know?

FlyWire produced one of the most detailed wiring maps of an adult brain in a complex animal. It is important infrastructure, not a complete explanation of brain function or behaviour.

When the wiring map is combined with real-time neural activity and behaviour experiments, which new predictions about learning, navigation or social behaviour become possible?

Return to the researchers' original publication

Nature research article and DOI FlyWire Connectome Data Explorer

DIJITALPI'S APPLICATION COMMENTARY

What could this look like in your organisation?

We created these scenarios to make the topic concrete. They are not cases from the paper or measured client results.

EXAMPLE 01

Fast classification, hidden error

A model labels thousands of images quickly, and the team reports only the total number processed.

Does processing more records prove that the map is correct?

Open the recommendation for example 01

Put an error measure beside speed.

Choose a balanced sample for manual review. Count false positives, false negatives and uncertain records separately. Repeat the same check when the model version changes.

EXAMPLE 02

A map is not an explanation of behaviour

After seeing a network visualisation, the team assumes it knows why the fly performs a behaviour.

Does a physical connection prove its function?

Open the recommendation for example 02

Separate structure from functional testing.

Use the wiring map to generate hypotheses. Test claims about behaviour or neural activity with experiments, recordings and comparisons.

EXAMPLE 03

Human review arrives too late

AI output enters the report, and experts inspect only a few examples before the presentation.

Does the timing of review affect quality?

Open the recommendation for example 03

Place review inside the workflow.

Define human approval at critical steps in advance. Keep the source record, review step and published finding connected in version history.

TRY IT WITH YOUR TEAM

Map the connections in one AI analysis.

Choose one AI analysis your team uses. Show its source data, automated classification, human review, corrections and published result in one flow. State separately which part of the flow was actually measured.

Explore our AI-powered working model

Source and ownership

Sven Dorkenwald, Arie Matsliah, Amy R. Sterling and the FlyWire Consortium.
FlyWire Consortium; Princeton University and partner research institutions.

Nature · 3 October 2024 · 634:124–138 · Peer-reviewed research. Open the original publication · Find on Google Scholar

The academic work belongs to the researchers named above. This page contains DijitalPi's explanatory review and original business examples; it is not a full translation of the paper.

← Return to all research articles