MACRO / REGION-SCALE CONNECTOMICS

The other scale.

A catalog of major public macro connectomics datasets: whole brains mapped between regions rather than between cells. Last reviewed August 2026. For synapse-resolution datasets, return to the main catalog.

MICRO

Cell to cell

Electron microscopy. Nodes are individual neurons, edges are synapses you can point at. Cubic millimetres at best, and every one is a decade of work.

MACRO

Region to region

Diffusion MRI or tracer injection. Nodes are anatomical areas, edges are estimated bundles. Whole living human brains, thousands of them, in an afternoon each.

These are not the same object at two zoom levels. A synapse-resolution connectome observes connections. Diffusion MRI infers them: tractography follows the direction water moves and reconstructs streamlines that are consistent with it, which is not the same as seeing an axon. Streamlines carry no direction, so a macro connectome is undirected unless a tracer supplied the polarity. Known failure modes include crossing fibres, gyral bias and a well-documented tendency to produce false positive bundles. The animal tracer atlases below sit in between: real projections with real polarity, at region scale.

Public macro datasets.

Human diffusion MRI cohorts first, then the animal tracer atlases that give them something to be checked against.

9records shown
Name / organismMethodScopeYearsAccess
Human Connectome Project, Young AdultHumanDiffusion MRI + resting and task fMRI1,206 healthy adults, 3T; a subset also at 7TThe reference dataset for human macro connectomics, and the source of most published human connectivity matrices.2010–2015Open, ConnectomeDB registration ↗UK Biobank ImagingHumanDiffusion MRI, resting and task fMRI, structuralTens of thousands imaged and still recruiting toward 100,000Imaging joined to deep phenotyping and genetics, which is what makes it different from a pure connectome resource.2014–ongoingApplication and fee ↗Adolescent Brain Cognitive DevelopmentHumanDiffusion MRI, resting and task fMRI11,880 children, longitudinal from ages 9–10The developmental counterpart: the same brains measured repeatedly rather than many brains once.2016–ongoingNIMH Data Archive, application ↗Developing Human Connectome ProjectHumanNeonatal and fetal diffusion MRI and fMRISeveral hundred neonates, plus a fetal cohortConnectivity before and around birth, where the tractography assumptions are hardest.2014–2022Open, registration ↗HCP Lifespan: Development and AgingHumanDiffusion MRI and fMRI, HCP-style protocolRoughly 1,300 aged 5–21 and 1,200 aged 36–100Extends the HCP protocol across the lifespan so the two ends are comparable to the young adult core.2016–2021Open, NDA / ConnectomeDB ↗Allen Mouse Brain Connectivity AtlasMouseAnterograde viral tracer, serial two-photon tomographyAround 2,900 injections mapped into one reference brainMesoscale, not MRI: real axonal projections between regions, which is a stronger claim than tractography can make.2014–Fully open ↗Marmoset Brain Connectivity AtlasMarmosetRetrograde tracer injections143 injections across cortex, cellular-resolution countsWeighted and directed, because retrograde tracing counts labelled cells rather than inferring a streamline.2016–Fully open ↗Brain/MINDS MarmosetMarmosetAnterograde tracer, MRI, histologyWhole marmoset brain, multimodalThe tracer and MRI layers are in one space, so tractography can be checked against ground truth in the same animal.2018–Open portal ↗CoCoMacMacaqueCollated published tract-tracingDecades of macaque tracing studies in one databaseNot one experiment but a curation of many, so coverage is uneven by design and the provenance matters.2001–Open ↗

Scope note. Two of these move. UK Biobank imaging is an open recruitment and ABCD is longitudinal with waves still being released, so both are given as released rather than as a total that would be wrong within months. Check the source before quoting a number.

Suggest a missing dataset ↗
SEE IT

What a macro connectome looks like.

Eighty-seven white matter bundles from the HCP group average, drawn inside the cortical surface they connect, with a signal running along one of them. This is the macro scale made visible: the highways, not the alleyways.

INTERACTIVE · HUMAN CONNECTOME PROJECT

The white matter, and a signal running through it

87 bundles from the HCP group average inside the group-average cortex. Drag to turn, and watch a pulse travel one tract at real relative length.

Open the visualisation ↗