How the Nervous System Assembles Itself, Day by Day express
Assembling a nervous system is like building a city where each new house, a neuron, must connect to the road network at the right time. In the tiny worm C. elegans, this process can be seen in detail: it has only 302 neurons, and scientists have reconstructed their contacts at different developmental stages.
The authors created DevoTG, a temporal neural network that learns from two types of data. The first is the cell lineage tree: which cell came from which and when. The second is eight snapshots of the worm's brain taken with an electron microscope on different days of its life. DevoTG doesn't just see a static picture; it remembers the sequence of events, like a foreman who watches the construction from the foundation pit to the finished building.
On the connection map, DevoTG identified three types of contacts. Some, stable, appear early and stay forever, like major highways. Others, temporary, are needed only during the growth phase, like scaffolding. The third type, variable, appear and disappear, like alleys that are repaved for new purposes. Especially interesting are the commander neurons AVA, AVB, and AVE: they take central positions early and only strengthen their role over time, accumulating connections.
This approach is not just an atlas but a movie about how the brain assembles itself. DevoTG's code is open, and it can be applied to other animals to reveal universal rules of neural construction.
Original abstract
as written by the authors arXiv:2606.21940
Understanding how a nervous system wires itself from birth to adulthood is a fundamental challenge in developmental neuroscience. We present DevoTG, a temporal graph framework that applies Temporal Graph Neural Networks (TGNs) to two complementary representations of C. elegans neural development: a Continuous-Time Dynamic Graph (CTDG) of cell division events derived from cell lineage data, and a Discrete-Time Dynamic Graph (DTDG) of the developing synaptic connectome spanning eight reconstructed electron-microscopy datasets. On the lineage prediction task, our TGN achieves a mean test AUC of 0.839 +/- 0.007 (5 seeds; validation AUC 0.937 +/- 0.001), outperforming a static GNN with the identical architecture by 26 AUC points (0.577 +/- 0.080), demonstrating that temporal memory is the decisive factor. Applied to the connectome DTDG, DevoTG identifies three connection stability classes (stable, developmental, and variable) across 225 neurons and 858 to 2,496 connections over development (L1 birth to adult), providing a temporal-graph-theoretic complement to the individual-variability classification of Witvliet et al. Analysis of hub command interneurons AVA, AVB, and AVE reveals their persistent centrality and how their integration roles are progressively reinforced across larval stages. Accompanying interactive visualizations (3D animated networks, centrality heatmaps, and a spatiotemporal lineage graph) make developmental dynamics accessible for biological hypothesis generation. DevoTG is open-source and designed for extension to other developing nervous systems. Code is publicly available at https://github.com/DevoLearn/DevoGraph/tree/main/DevoTG.
The worm C. elegans has 302 neurons, and scientists know the complete wiring diagram of its connections: it's the only creature for which such a map has been made.