Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Jun;642(8067):398-410.
doi: 10.1038/s41586-025-08985-1. Epub 2025 May 7.

Light-microscopy-based connectomic reconstruction of mammalian brain tissue

Mojtaba R Tavakoli  1 , Julia Lyudchik  1 , Michał Januszewski  2 , Vitali Vistunou  1 , Nathalie Agudelo Dueñas  1 , Jakob Vorlaufer  1 , Christoph Sommer  1 , Caroline Kreuzinger  1 , Bárbara Oliveira  1 , Alban Cenameri  1 , Gaia Novarino  1 , Viren Jain  3 , Johann G Danzl  4
Affiliations

Light-microscopy-based connectomic reconstruction of mammalian brain tissue

Mojtaba R Tavakoli et al. Nature. 2025 Jun.

Abstract

The information-processing capability of the brain's cellular network depends on the physical wiring pattern between neurons and their molecular and functional characteristics. Mapping neurons and resolving their individual synaptic connections can be achieved by volumetric imaging at nanoscale resolution1,2 with dense cellular labelling. Light microscopy is uniquely positioned to visualize specific molecules, but dense, synapse-level circuit reconstruction by light microscopy has been out of reach, owing to limitations in resolution, contrast and volumetric imaging capability. Here we describe light-microscopy-based connectomics (LICONN). We integrated specifically engineered hydrogel embedding and expansion with comprehensive deep-learning-based segmentation and analysis of connectivity, thereby directly incorporating molecular information into synapse-level reconstructions of brain tissue. LICONN will allow synapse-level phenotyping of brain tissue in biological experiments in a readily adoptable manner.

PubMed Disclaimer

Conflict of interest statement

Competing interests: ISTA filed a patent application covering the expansion technology, with M.R.T. and J.G.D. as inventors (application number EP24157341.9). M.J. and V.J. are employed by Google, which sells cloud computing services. The remaining authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Dense connectomic reconstruction of mammalian brain tissue with light microscopy.
a, LICONN volume of around 1 × 106 μm3 (native tissue scale) of mouse primary somatosensory cortex (layers II/III–IV, 396 × 109 × 22 µm3 original tissue scale, 0.95 × 106 µm3 before and 3.5 × 109 µm3 after hydrogel expansion of approximately 16×). Seventy-nine example cells from dense reconstruction with FFN. Right, dendrite from pyramidal neuron (box at top of left panel), with deep-learning predictions of the synaptic molecules bassoon (cyan, pre-synapses) and SHANK2 (magenta, excitatory post-synapses) and synaptically connected axons (bottom). Scale bars, 15 μm (left); 2 μm (right). Length scales and scale bars refer to biological size before expansion throughout. b, Subregion (single plane) of a. Enlarged views: top, intracellular structure of pyramidal neuron; bottom, primary cilium. Spinning-disk confocal imaging data with contrast-limited adaptive histogram equalization (CLAHE) (for comparison of raw versus CLAHE, see Supplementary Fig. 12). Scale bars, 5 μm (left); 2 μm (top right); 1 μm (bottom right). c, Single plane (with CLAHE) of dendrite with spine (box) and cell nucleus with nuclear pores (bright densities, bottom). Small panels: top, excitatory synapse with presynaptic protein-rich punctate features and bar-like feature at post-synapse (without CLAHE). Line: direction of intensity profiles (top right), measuring distance between pre- and postsynaptic features (DP: dense projection; mean ± s.d.) with violin plot. r: coordinate along line profile. Middle, single bouton contacting two spines and DP–DP distance (mean ± s.d.). Bottom, highlighted spine. Scale bars, 2 μm (main image); 100 nm (right images). d, Periodic, protein-dense structure at circumference of neurite subset (without CLAHE). Example line profile and periodicity (mean ± s.d.). Scale bars, 1 μm. e, Manual axon tracing. Top left, single plane (magnified) of LICONN volume (around 19 × 19 × 19 µm3) from primary somatosensory cortex in a Thy1-eGFP mouse with cytosolic eGFP expression in an axon (arrow). Top right, overlay with immunolabelling for eGFP. Scale bars, 1 μm. Bottom, renderings (green) and skeletons (black) of eight eGFP-expressing axons (ground truth, based on eGFP and structural LICONN channels), and skeletons generated by two annotators blinded to eGFP signal (magenta, consensus, offset for clarity). For additional datasets, see Supplementary Figs. 13 and 14. Scale bar, 5 μm. f, Manual dendrite tracing. Top, single plane (magnified) from LICONN volume (around 19 × 19 × 19 µm3, hippocampus, CA1; Thy1-eGFP mouse) with eGFP-expressing dendrite (arrow) and overlay with eGFP (green). Middle, cross-sections of eGFP-expressing dendrite with spines. Scale bars, 1 μm. Bottom, dendrite skeleton (black) generated from eGFP and structural channels (ground truth), within 3D rendering (green). Additional skeleton (magenta) generated from structural channel by two annotators blinded to eGFP. For additional datasets, see Supplementary Fig. 15. Scale bar, 5 μm. g, Left, single-tile LICONN volume (hippocampus, CA1) with manual cellular annotations (colour, 658 structures) and 3D rendering. Middle, top view. Right, neurites and magnified synaptic connections (different camera position). Scale bars, 5 μm (left); 3 μm (middle and right). Source Data
Fig. 2
Fig. 2. Deep-learning-based segmentation.
a, Rendering of 85 × 69 × 14-µm3 LICONN volume (native tissue scale) from hippocampal neuropil in CA1, overlaid with dense FFN-based segmentation of neuronal structures in the bottom corner. Neuronal structures were comprehensively proofread in this volume by correction of split and merge errors (without any manual painting of voxels; see https://neuroglancer-demo.appspot.com/#!gs://liconn-public/ng_states/expid82.json for original data and proofread segmentation). Scale bar, 10 μm. b, Magnified view from a single plane with (top to bottom) raw structural data (CLAHE applied), dense segmentation after proofreading and overlay. Scale bars, 1 μm. c, Rendering of 5.8% of axons contained in the volume in a (see https://neuroglancer-demo.appspot.com/#!gs://liconn-public/ng_states/expid82_fig2_axons.json for browsable data). Scale bar, 10 μm. d, Rendering of 27.3% of dendrites and a small number of axons (see https://neuroglancer-demo.appspot.com/#!gs://liconn-public/ng_states/expid82_fig2_dends.json for browsable data). e, Rendering of example dendrites. Scale bar, 10 μm. f, Spatial arrangement of selected axonal and dendritic structures, highlighting various types of contact. Scale bar (top right), 2 μm. g, Segment size (number of voxels on a logarithmic scale) of neuronal structures for the base FFN segmentation (blue), after automated agglomeration (white) and after full manual proofreading of the automated agglomeration (yellow) for the dataset in a (n = 1 dataset). Vertical bar, lower and upper quartiles; dot, median; vertical lines, 1.5× interquartile range. h, Edge accuracy for the base segmentation (blue), after automated agglomeration (white) and after manual proofreading (yellow). i, Distribution of spine-head volumes in the same segmentation volume, analysed for 59,332 spines. Percentage numbers refer to the intervals indicated by the vertical lines (<0.01, 0.01–0.05, 0.05–0.1, 0.1–0.2, >0.2 µm3). Source Data
Fig. 3
Fig. 3. Molecular labelling and synapse detection.
a, Immunolabelling for presynaptic bassoon (cyan) and postsynaptic PSD95 (magenta) in LICONN volume (somatosensory cortex). Right, magnified views of synapses (boxes in left panel) showing immunolabelling and structural channels separately and overlaid. Scale bars, 2 μm (left); 500 nm (right). b, Distance between bassoon and SHANK2 signals (mean ± s.d., violin plot including median and quartiles, 106 synapses). c, Three-dimensional renderings of bassoon and SHANK2 immunolabelling mapped onto a dendrite from FFN segmentation of the volume in a. Scale bar, 2 μm. d, Top, LICONN with immunolabelling for synaptic markers. Bottom, overlay with structural channel. Single planes from volumes in hippocampal CA3 (stratum lucidum, leftmost panel) and CA1. Scale bars, 500 nm. e, Left, dendritic spine with SHANK2 (magenta) and GLUN1 (cyan) immunolabelling, showing maximum intensity projections of respective immunolabellings (points: centres of mass). Scale bars, 100 nm. Right, mean ± s.d. and violin plot of centre-of-mass distances, including median and quartiles (184 synapses). f, Illustration of excitatory-synapse detection through bassoon (cyan) and SHANK2 (magenta) immunolabellings, converted to point annotations. Examples include 1:1 and 1:2 presynaptic to postsynaptic connections. Scale bar, 100 nm. g, Immunolabelling renderings with detected synapses (2:1, 1:1 and 1:2 presynaptic to postsynaptic connections). Cyan and magenta balls, pre- and post-synapses; grey and black bars, computationally detected connections and manually generated ground truth. h, Ground-truth (proofread) immunolabelling-based excitatory-synapse detections in a 913-µm3 volume (hippocampus, CA1, stratum radiatum). Scale bar, 2 μm. i, Three-dimensional rendering of dendrite (hippocampus, FFN segmentation) with excitatory synapses (bars), detected through bassoon (pre-synapses, cyan) and SHANK2 (post-synapses, red) immunolabelling. Magnified views include synaptically connected boutons. Scale bar, 10 μm. j, LICONN with immunolabelling for gephyrin (yellow, inhibitory post-synapses) and SHANK2 (magenta, excitatory post-synapses), with immunolabelling shown separately and overlaid for the boxed region. Inhibitory post-synapses show less pronounced structure than excitatory post-synapses. Scale bars, 2 μm (left); 500 nm (right). k, Left, LICONN with immunolabelling for VGAT (cyan, inhibitory pre-synapses) and gephyrin (yellow), with channels shown separately for the boxed region. Right, similar measurement with immunolabelling for bassoon (cyan, excitatory and inhibitory pre-synapses) and gephyrin (yellow). Scale bars, 1 μm (main); 500 nm (enlarged boxes). Source Data
Fig. 4
Fig. 4. Connectivity analysis and deep-learning-based synapse detection.
a, Spiny dendrite with SHANK2-expressing excitatory (magenta) and gephyrin-expressing inhibitory post-synapses (yellow), connected axons (presynaptic bassoon: cyan) and seed locations. b, Input density onto spiny dendrites (primary somatosensory cortex), defined by immunolabelling: excitatory (SHANK2+); inhibitory (gephyrin+); IF+: total immunofluorescence. Numerical values: mean ± s.d. throughout; box plots: median; lower and upper quartiles; whiskers: minimum and maximum (throughout). Data points: individual dendrites (11; 123 µm total, 351 IF+ synapses, b–f). c, Synapse (SHANK2+ or gephyrin+) target locations. d, Molecular properties of spine heads, with synapses identified by immunolabelling or PSDs. e, Molecular properties of shaft synapses. f, Excitatory (E) versus inhibitory (I) synapses onto dendrites. g, Spine seeding by excitatory axon outputs. h, Output density of spine-seeded axons (19 axons (data points) with two or more outputs, total length 990 µm). i, Spine fraction traced to parent dendrite for spines contacted by the same 19 axons (128 spines, mean ± s.d. over individual axon). Data points: axons. j, Inhibitory axons seeded at an AIS (gephyrin+, no PSD). k, Output fraction onto shafts of AIS-seeded axons (eight axons with three or more outputs). l, Deep-learning prediction of bassoon and SHANK2 location from structural LICONN channel. Single plane from LICONN volume (CA1, stratum radiatum, not included in training), comparing prediction with immunolabelling. Scale bars, 1 μm. m, Corresponding volumetric renderings. n, Excitatory input and output for a pyramidal neuron from the dataset in Fig. 1a. Synapse detections through bassoon and SHANK2 prediction mapped onto FFN segmentation. Magenta numbers: detected synapses between indicated branch points. Magnified views: structural channel, molecule predictions and cellular segments (partial proofreading, eliminating false-positive detections without adding missed detections). Scale bars, 2 μm (top); 20 μm (middle); 500 nm (bottom). o, Connectivity prediction in hippocampal dataset from Fig. 2a. Left, rendered synapse predictions. Middle, axon (black) with connected dendrites. Right, dendrite (black) with connected axons. Scale bar, 10 μm. p, Integration of structural, immunolabelling and deep-learning analysis. Dendrite (primary somatosensory cortex) with immunolabelling-based detection of excitatory (501) and inhibitory (80) post-synapses. Insets: presynaptic partners identified by deep-learning bassoon prediction. Scale bar, 10 μm. Source Data
Fig. 5
Fig. 5. Applications of LICONN beyond synaptic connectivity.
a, Molecular identification of interneurons through somatostatin immunolabelling (SST+, cyan). Overview plane of LICONN volume (cortex), with maximum intensity projection of immunolabelling. Nuclear infoldings: orange. Scale bar, 5 μm. b, Ankyrin G immunolabelling (red) (cortex), with periodic protein-density modulation in structural channel, highlighting the AIS. Magnified images: channels shown separately. Low-protein-density voids (straight arrows) around axons indicate myelination (Extended Data Fig. 5). Scale bars, 1 μm. c,d, Primary cilia in LICONN (cortex, hippocampus; c) with intensely labelled centrioles (schematic; d) in the basal body. Scale bars, 500 nm. e, Immunolabelling for acetylated tubulin (magenta) and adenylate cyclase 3 (cyan) at a primary cilium (hippocampus, CA1) and overlay with the structural LICONN channel, with the membrane-bound adenylate cyclase signal ensheathing axoneme. Scale bars, 500 nm. f, Length of primary cilia (mean ± s.d. throughout) according to cell type in dataset from Fig. 1a (78 cells; box plots: median; lower and upper quartiles; whiskers: minimum and maximum (throughout); points: individual cells). g, Length of primary cilia in wild-type and Hnrnpu+/− mice (hippocampus, CA1, pyramidal layer; Supplementary Fig. 25). h, LICONN volume (border of corpus callosum and alveus; Extended Data Fig. 9) with immunolabelling for acetylated tubulin (red), revealing multi-ciliated cells. Scale bar, 1 μm. i, Cross-section of cilium with ninefold symmetry in protein density, probably reflecting microtubule doublets, with ring diameter and doublet distance. Scale bar, 100 nm. j, GFAP (red, astrocytes) and glutamate receptor GLUN1 (magenta) immunolabelling with deep-learning bassoon prediction (cyan) in LICONN volume (hippocampus). Magnified images: astrocytic primary cilium at different z-planes, apposed to synaptic boutons. Scale bars, 1 μm (main image); 500 nm (right images). k, Immunolabelling for the astrocytic gap-junction protein connexin-43 (orange) and the inhibitory-synapse marker gephyrin (yellow) in LICONN (cortex). Magnified images, gap junction between astrocytes. Scale bars, 1 μm (main image); 500 nm (right images). l, ‘Virtual’ five-colour measurement (cortex). LICONN with connexin-43 (orange) and (gephyrin, yellow) immunolabelling, deep-learning prediction of pre-synapses (bassoon, cyan) and excitatory post-synapses (SHANK2, magenta). Channels shown separately for boxed region. Scale bars, 1 μm. m, Density of gephyrin-positive inhibitory synapses and connexin-43-positive gap junctions (four volumes; cortex, hippocampal CA1). Source Data
Extended Data Fig. 1
Extended Data Fig. 1. Tracing of neuronal structures in the cortex.
Manually generated skeletons in the dataset in Fig. 1a, showing traceability of axons and dendrites across borders of fused imaging tiles. Magnified views indicate examples of glial cells, primary cilia used in further analysis (Fig. 5) and dendrites. Spines of a particular dendrite were only comprehensively traced in the right part of the dendrite indicated in the magnified panel ii.
Extended Data Fig. 2
Extended Data Fig. 2. Hippocampal architecture analysed with LICONN.
Hippocampal overview composed of 2,969 high-resolution tiles from spinning-disc confocal microscopy with delineation of subregions and layers. Yellow boxes indicate the regions magnified in the bottom panels.
Extended Data Fig. 3
Extended Data Fig. 3. Synaptic labelling in mossy fibre boutons in the CA3 stratum lucidum.
a, Mossy fibre boutons forming synapses with complex spines (thorny excrescences) at proximal dendrites of CA3 pyramidal neurons in hippocampus, interspersed with bundles of mossy fibres (non-myelinated axons of DG granule cells). Single plane of a LICONN imaging volume including the structural channel (grey) and immunolabelling for VGLUT1 (vesicular glutamate transporter 1, magenta), a synaptic vesicle marker in excitatory synapses, and the active-zone marker RIM1/2 (cyan). Bottom: Immunolabelling and structural channels displayed separately and as overlay for the two mossy fibre boutons and their postsynaptic partners in the box in the top panel. b, Similar measurement with immunolabelling for presynaptic bassoon and the postsynaptic scaffolding protein PSD95, present at excitatory synapses. Immunolabelling for RIM1/2 and VGLUT1 is representative of n = 2 technical replicates and bassoon/PSD95 of n = 3 replicates across n = 2 mice.
Extended Data Fig. 4
Extended Data Fig. 4. Excitatory and inhibitory synaptic inputs mapped onto reconstructed cell somata.
Reconstruction of four example cell somata from LICONN datasets in somatosensory cortex with immunolabelling for SHANK2 and gephyrin. Excitatory (magenta) and inhibitory (yellow) synaptic inputs are mapped onto cell body reconstructions. Synapses were classified as excitatory if (1) a typical presynaptic structure, (2) a PSD and (3) SHANK2 were present. Synapses were classified as inhibitory if (1) a presynaptic structure was present and (2) a PSD was missing, with or without gephyrin. For each cell body, the number of excitatory and inhibitory connections are indicated, showing a much higher number of inhibitory inputs. The one cell on the right displayed a reversed pattern with 30 SHANK2-positive synapses. We analysed 13 cell bodies (contained ≥50% within the imaging volume) across 8 datasets and counted 48.7 ± 19.8 (mean ± s.d.) inhibitory synapses per soma (excluding the one cell with reversed input preference displayed on the right). Excitatory synapses were sparse, with most somata receiving 0 and one soma receiving 1 excitatory input.
Extended Data Fig. 5
Extended Data Fig. 5. Molecular identification of cellular and subcellular structures.
a, Single plane of a LICONN imaging volume in the hippocampal CA1 stratum radiatum, including the structural channel (grey) and immunolabelling for myelin basic protein (magenta) and the astrocytic gap-junction protein connexin-43 (cyan). Magnified views highlight structures expressing the respective molecular markers: (i) Immunolabelling for myelin basic protein identifies myelin sheaths, where the regions largely devoid of signal in the structural LICONN channel correspond to lipid-rich myelination. (ii) Connexin-43 indicates gap junctions. b, Single plane from LICONN imaging volume in hippocampus, showing the structural channel (grey) and immunolabelling for GFAP (orange) and connexin-43 (cyan). GFAP is an intermediate filament protein, commonly used as astrocyte marker. Note the GFAP signal in the immediate vicinity of the capillary traversing the image, and the cell nucleus on the right in the periphery of the blood vessel. Panels in the bottom show the immunolabelling and structural channels separately, as indicated by the box. c, Single plane from a LICONN imaging volume in the CA3 stratum lucidum, with additional immunolabelling for vimentin (orange, here highlighting the endothelial lining of a blood capillary) and the peroxisomal marker PMP70 (70 kDa peroxisomal membrane protein). Note that peroxisomes do not produce pronounced contrast in the structural (protein density) channel but can be identified by specific labelling. The displayed combinations of immunolabellings for MBP/Cnx-43 were technically replicated n = 1 time, RIM1/2 and VGLUT1 n = 2 times, GFAP/Cnx-43 n = 2 times, vimentin/PMP70 n = 1 time.
Extended Data Fig. 6
Extended Data Fig. 6. LICONN analysis in diverse brain regions.
Single example planes from LICONN imaging volumes in hypothalamus, piriform area and cerebellum. In the cerebellum, additional immunolabelling for bassoon and SHANK2 was applied.
Extended Data Fig. 7
Extended Data Fig. 7. LICONN imaging in the hippocampal CA3 stratum lucidum.
Single-plane overview image with magnified views, illustrating the complex arrangement of cellular structures including a primary cilium and intracellular organelle diversity in (i) and mossy fibre bundles and an example of multiple shaft synapses in (ii). Overview imaging in CA3 stratum lucidum was performed in n = 2 technical replicates, single tile measurements were performed in multiple technical replicates.
Extended Data Fig. 8
Extended Data Fig. 8. LICONN imaging in white matter.
a, Schematic of a mouse coronal section and the region of imaging (box). b, Single planes from individual LICONN imaging volumes spanning deep cortex, corpus callosum, alveus and hippocampus according to the schematic, revealing the distinct organization in the various regions. Imaging of the various layers was performed in n = 1 mouse.
Extended Data Fig. 9
Extended Data Fig. 9. Multi-ciliated cells at the transition between the corpus callosum and the alveus.
a, Regional overview image. b, 3D rendering of a LICONN imaging volume (top, recorded with a 20× objective lens) of the region indicated by the box in the overview, highlighting multi-ciliated cells. The bright structures correspond to basal bodies. Bottom: Single imaging planes from the same volume as indicated by the magenta box in the upper panel. Intensity lookup tables in the individual panels are adjusted to either emphasize the intensely labelled basal bodies or the local cellular architecture. c, Left: Rendering of a LICONN imaging volume recorded with a high-NA objective lens in the same region from a different specimen. Right: Single imaging plane showing multi-ciliated cells and magnified view of cilia with intensity lookup table adjusted to visualize the axonemes of cilia. Representative of replicates in n = 2 mice.
Extended Data Fig. 10
Extended Data Fig. 10. Scaling up in axial direction through iterative imaging and sectioning of LICONN tissue–hydrogel hybrid.
a, LICONN in 300 µm-thick brain slices (hippocampus, CA1, stratum radiatum). Single imaging planes in subregions of multi-tile volume, with data quality equivalent to thinner tissue sections. Boxed regions: magnified views. b, Schematic of iterative block-face imaging/sectioning of LICONN tissue–hydrogel hybrid. In round 1, a multi-tile volume is acquired, then most of the imaged hydrogel layer is removed with a vibratome, and in round 2 a second multi-tile imaging volume deeper in the tissue is acquired. A region of continuous overlap recorded in both imaging rounds enables lossless stitching and fusion of imaging volumes across rounds. c, Axial (xz-)view of the LICONN volume of panel a, fused from 72 subvolumes arranged on a 6x6x2 3D grid. Data were acquired in two imaging rounds at increasing depth (data displayed after coarse stitching). Axial extent, round 1: 466 µm (30.2 µm native tissue scale) with vibratome cut 320 µm below hydrogel surface. Axial extent, round 2: 353 µm (23 µm native scale). Total multi-round acquisition time: 3 h 36 min (imaging round 1: 2 h 04 min, sample handling and vibratome cutting: 15 min, imaging round 2: 1 h 17 min). d, Single-plane overview of the same volume in lateral (xy-)direction (109x109 µm2 native scale) after volume fusion. Magnified view (boxed region): region with tile borders, showing seamless fusion (region adjacent to view in left image in panel a). e, Single-plane axial (xz-)view after volume fusion in the same dataset, with seamless fusion across the two imaging rounds. f, Single-plane lateral (xy-)view of a subarea in the axial overlap region between imaging rounds, after volume fusion. Data from rounds 1 and 2 are colour coded in red and green. Homogeneous colouring indicates voxel-exact matching from individual imaging rounds after fusion. g, 3D-illustration of manually traced axons in the final stitched and fused volume, demonstrating traceability across borders between imaging rounds. h, Single axial plane of LICONN volume fused from 108 subvolumes arranged on 3x3x12 grid in 3D, acquired in 12 consecutive block-face imaging/sectioning rounds, and 3D-illustration of manually traced axons. Axons extend across tile borders in 3D, from axon entering to axon leaving the imaging volume. Axial extent of final fused volume: 205 µm original tissue scale (7926 optical sections, 400 nm z-steps in post-expansion volume). Original data (CLAHE-processed) of final fused volume: https://neuroglancer-demo.appspot.com/#!gs://liconn-public/ng_states/multiround_fusion.json.
See this image and copyright information in PMC

References

    1. Kasthuri, N. et al. Saturated reconstruction of a volume of neocortex. Cell162, 648–661 (2015). - DOI - PubMed
    1. Jefferis, G., Collinson, L., Bosch, C., Costa, M. & Schlegel, P. Scaling up Connectomics: The Road to a Whole Mouse Brain Connectome. https://wellcome.org/reports/scaling-connectomics (Wellcome Trust, 2023).
    1. Kornfeld, J. & Denk, W. Progress and remaining challenges in high-throughput volume electron microscopy. Curr. Opin. Neurobiol.50, 261–267 (2018). - DOI - PubMed
    1. White, J. G., Southgate, E., Thomson, J. N. & Brenner, S. The structure of the nervous system of the nematode Caenorhabditis elegans. Phil. Trans. R. Soc. B314, 1–340 (1986). - PubMed
    1. Scheffer, L. K. et al. A connectome and analysis of the adult Drosophila central brain. eLife9, e57443 (2020). - DOI - PMC - PubMed

LinkOut - more resources