Cell morphology · representative shape

Click a neuron in the network to see a representative reconstructed morphology of its cell type (from NeuroMorpho.org).
Morphologies: representative published reconstructions from NeuroMorpho.org (Ascoli et al., 2007) — one per cell type; IDs & PMIDs in CA1_morphology_sources.md. Representative of the type, not the modelled cell.
— · fps
Drag to rotate · scroll to zoom · click a cell to trace it · hover for info

Human CA1 — Connectome

Single-cell wiring of the human hippocampal CA1, on your GPU. Cells coloured by type; the tour walks the key findings.
Loading the human connectome…
Load local .bin files instead
ⓘ About this dataset
Region. Hippocampal CA1 — one stage of the trisynaptic circuit (entorhinal cortex → dentate gyrus → CA3 → CA1 → subiculum → entorhinal cortex). CA1's main input is the CA3 Schaffer collaterals; it also gets direct entorhinal input and projects to the subiculum. Here only CA1 cells are modelled — CA3/entorhinal afferents enter as external drive.

Model. Full-scale human CA1 (Gandolfi, Mapelli, Solinas & Migliore; EBRAINS / EU Human Brain Project). The complete model is ~5.28 million neurons (4.8M pyramidal + 0.48M interneurons, ~9% inhibitory). This explorer shows all 5.28M neurons; the full connectome is ~41 billion synapses, so wiring is a representative central-column subset (cells outside that column appear as points without wiring).

Neurons. PYR (pyramidal) are the excitatory principal cells — ~90% of CA1; they drive their targets and are the output of the region. Every other type is a GABAergic interneuron (NGF, Ivy, and others) that locally inhibits pyramidal cells. Cells are coloured by type (see legend).

Connections. Each "Findings tour" entry is one pathway: a source cell type wiring to a target type (e.g. NGF→PYR). A wire is coloured by the cell it comes from. Click any cell to trace its outgoing connections, broken down by target type. Click the 🧬 morphology button, then a cell, to see a representative reconstruction of its type.

Morphologies. The scaffold model stores only soma positions (no shapes). For visualization we attach one representative published reconstruction per cell type from NeuroMorpho.org (Ascoli et al., 2007; CA1, mostly mouse). These are representative of the type — not the individual modelled cell — and carry no quantitative claim. Per-type NeuroMorpho IDs and PMIDs are listed in CA1_morphology_sources.md.

What we found. Neurogliaform (NGF) cells form a giant inhibitory blanket over the pyramidal population — in this model via distance-independent connectivity, so a single NGF cell reaches clear across CA1 while every other interneuron class stays local. NGF also receives no pyramidal input in the model, so pyramidal activity cannot recruit it. Ivy forms a second dendritic blanket but, unlike NGF, receives pyramidal drive — a closed feedback loop. Select the NGF blanket and then Ivy blanket in the findings tour: the difference in wire length is the whole point, and it is visible without any statistics. Quantification is in preparation.

Activity. Not included in this public build. The spike data is an unpublished simulation run and is not ours to publish; the wiring shown here is derived from the published Gandolfi et al. (2023) model. If you have the spike file locally, "Load local .bin files instead" will play it.
Drag rotate · right-drag pan · wheel zoom · click a cell → its targets.
Keys: 1–9, 0, − tour stops (in dropdown order) · H hubs · C clear · Space play/pause · R reset view.
Developed by Safeer Ali Mirani, PhD
University of Sassari · safeer.ali.mirani@gmail.com
safeeralimirani.netlify.app ↗
Full-scale CA1 model — Gandolfi, Mapelli, Solinas & Migliore (EBRAINS / EU Human Brain Project); simulated on CINECA Galileo100. Please cite the source model when reusing.
build v3 · filters · timeline · find · rates · video