Synister

(Eckstein et al., 2024)

High-resolution electron microscopy of nervous systems has enabled the reconstruction of synaptic connectomes. However, we do not know the synaptic sign for each connection (i.e., whether a connection is excitatory or inhibitory), which is implied by the released neurotransmitter.

We demonstrate that artificial neural networks can predict transmitter types for presynapses from electron micrographs: a network trained to predict six transmitters (ACh, Glut, GABA, Ser, Dop, Oct) achieves an accuracy of 87% for individual synapses and 94% for entire neurons.

We visualize the ultrastructural features used for prediction, discovering subtle but significant differences between neurotransmitter phenotypes.

We also analyze neurotransmitter distributions across the brain and find that neurons that develop together largely express only one fast-acting neurotransmitter (ACh, Glut, or GABA).

We hope that our publicly available predictions act as an accelerant for neuroscientific hypothesis generation for the fly.