Exploring neurodevelopment via spatiotemporal collation of anatomical networks with NeuroSC
- Department of Neuroscience and Department of Cell Biology, Wu Tsai Institute, Yale University, United States
- Department of Genetics, Yale School of Medicine, United States
- Department of Biology, Brigham Young University-Idaho, United States
- Bilte Co., United States
- Program for Applied Mathematics, Yale University, United States
- Department of Computer Science, Yale University, United States
- Program for Computational Biology and Bioinformatics, Yale University, United States
- Department of Genetics and Genome Sciences and Center for Cell Analysis and Modeling, University of Connecticut Health Center, United States
- MBL Fellow, Marine Biological Laboratory, United States
- Wu Tsai Institute, Yale University, United States
- Instituto de Neurobiología, Recinto de Ciencias Médicas, Universidad de Puerto Rico, Puerto Rico
Figures
Figure 1
DC/C-PHATE representations of contactome-based relationships.
DC/C PHATE graphs enable representations of neuronal contact relationships. To build DC/C-PHATE graphs, we (A) analyzed serial section EM datasets of the C. elegans nerve ring neuropil (located in the head of the animal). (B) Single cross-section of the nerve ring (surrounding the pharynx), with segmented neurites pseudo-colored. The dark box corresponds to the zoomed-in image in (C). The cross-section is from the JSH dataset digitally segmented (Brittin et al., 2021). (C) Zoom-in cross section with three arbitrary neurons (called A, B, C) highlighted by overlaying opaque cartoon (2-D, left image) and 3-D shapes (middle image) to represent the segmentation process in the z-axis (arrow) and the neuronal contact sites (highlighted Yellow, orange, and Red). Contacts are quantified for all neuron pairs across the contactome (see Materials and methods), to generate a Contact Matrix (represented here as a table, schematized for the three arbitrary neurons selected and in which specific contact quantities are represented by a color scale and not numerical values). Yellow represents little contact, and red represents a large degree of contact. Here, as an example, you can see that neuron B and C have the largest degree of contact. In an actual contact matrix, this would be a large number of shared pixels. (D) Schematic of how the Diffusion Condensation algorithm (visualized with C-PHATE) works. DC/C-PHATE makes use of the contact matrix to group neurons based on similar adjacency profiles (Brugnone et al., 2019; Brugnone et al., 2019; Moyle et al., 2021), schematized here for the three neurons in (C). (E) Screenshot of the 3-D C-PHATE graph from a Larval stage 1 (L1; 0 hours post hatching;) contactome, with individual neurons represented as spheres at the periphery. Neurons were iteratively clustered towards the center, with the final iteration containing the nerve ring represented as a sphere in the center of the graph (Highlighted in maroon). (F) Integration in NeuroSC of the DC/C-PHATE and EM-derived 3-D neuron morphology representations allows users to point to each sphere in the graph and determine cellular or cluster identities for each iteration. Shown here and circled in red, an arbitrarily selected cluster (in E), with the identities of the neurons belonging to that cluster (four letter codes in the column to the left of F) and the corresponding neuronal morphologies (right) of this group of neurons in the EM-reconstructed nerve ring (with individual neurons pseudo-colored according to their names to the left). Compass: Anterior (A), Posterior (P), Dorsal (D), Ventral (V), Left (L), Right (R).
Figure 2
Implementation of DC/C-PHATE to developmental contactomes reveals a conserved layered organization maintained during post-embryonic growth.
(A) Cartoon of the C. elegans head and nerve ring (outlined with black box). Below, nerve ring reconstruction from EM data of an L1 animal (5 hours post-hatching), with all neurons in gray. Scale bar 2 μm. (B–F) DC/C-PHATE plots generated for available contactomes across C. elegans larval development, colored by stratum identity as described (Moyle et al., 2021). Individual neurons are located at the edges of the graph and condense centrally. The four superclusters identified and all iterations before are colored accordingly. The identity of the individual neurons belonging to each stratum, and at each larval stage, was largely preserved and is provided in Supplementary file 3; Supplementary file 4; Supplementary file 5; Supplementary file 6. Some datasets contain 5 or 6 super-clusters (colored hues of the stratum that they most closely identify with). These clusters are classified as groups of neurons that are differentially categorized across the developmental connectomes. Note in B the blue cluster extends far to the left due to rotation of the 3D image. (G–K) Volumetric reconstruction of the C. elegans neuropil from EM serial sections for the indicated larval stages (columns) with the neurons colored based on their strata identity. Scale bar 2 μm; Anterior (A) left, Dorsal (D) up.
Figure 3
Examination of the architectural motifs underlying the distinct strata across development.
Visualization of (A–F) Stratum 1 (Red) and (G–L) Strata 3 and 4 (Blue and Green) reveal motifs that are preserved (Stratum 1) and change (Strata 3 and 4) across developmental contactomes (L1 to Adult, left to right, as indicated by labels on top). (B–F) Cropped view of Stratum 1 at each developmental stage showing a similar shape of two ‘horn-like’ clusters in the C-PHATE graphs (as seen by orange and blue shaded areas). These two clusters have similar neuronal memberships, which are largely invariant across developmental contactomes (Supplementary file 3). (H–L) Cropped view of Strata 3 and 4 at each developmental stage highlighting differences in the organization and number of neurons contained in each of the Blue and Green strata, which is particularly distinct when comparing (H) L1 and (K) L4 (Supplementary file 5; Supplementary file 6).
Figure 4 with 2 supplements
Case study: AIML and PVQL neurons change clustering patterns across the developmental contactomes.
(A–E) C-PHATE plots across development, with the trajectories of AIM neurons (in purple) and the rest of the spheres colored by stratum identity (see Figure 2). (F–G) Zoom in of the AIML and PVQL trajectories corresponding to larval Stage 1 (pre-AVF ingrowth) (A, dotted box) and in (G), Larval Stage 3 with AVFL/R present (C, dashed box). Note how the relationship between AIM and PVQ neurons in the C-PHATE graph varies for each of the examined contactomes across development. (Figure 4—figure supplement 1, Supplementary file 7). (H,I) simplified schematics of F and G based on neuron class.
Figure 4—figure supplement 1
DC/C-PHATE clustering of AIM, PVQ, and AVF across postembryonic development.
(A–E) A cropped view of the DC/C-PHATE plot colored to identify individual neurons and clustering events in (A) Larval stage 1 (5 hours post hatching); (B) Larval stage 2 (23 hours post hatching); (C) Larval Stage 3 (27 hours post hatching); (D) Larval stage 4 (36 hours post hatching); and (E) Adult (48 hours post hatching). See also Figure 4—video 1 and Supplementary file 7.
Figure 4—video 1
Visualization of hierarchical relationships using C-PHATE plots in NeuroSC.
The process for rendering a C-PHATE plot at the L4 stage (36 hours post hatching) with the real-time loading speed. In the viewer, 3-D visualization of a C-PHATE plot (shades of cyan), which is rotated to show the dome- shape of the plots and to orient the plot to correspond to Figures 2 and 4. The highlight functionality is used to show the spheres containing AIM (teal), then PVQ (teal). The spheres of the first iterations, containing AIM and PVQ, are identified, selected, and colored magenta. The AVF neurons are highlighted in teal, and the first AIM and AVF containing clusters are identified, selected, and colored yellow. The first clusters containing AIML, AVF, and PVQL are identified and colored green. Neurons in the left yellow and magenta clusters are reconstructed with a right click on the sphere and ‘“Add to new viewer’” selection.
Figure 5 with 2 supplements
Case Study: Visualization of contact profiles in individual neurons.
(A) Cartoon schematic of the head of the animal with the AIM neurons (purple) and pharynx (gray), and (dotted box) a 3-D reconstruction of the AIM neuron morphology from the L1 (0 hours post-hatching) dataset. (B) AIML and AIMR neurites rendered in 3D from L1. Note that we did not implement any surface smoothing methods to objects, so there might be gaps in the renderings. This was done intentionally, with the goal of producing the most accurate representation of the available data segmentation and avoiding any rendering interpretations. (C) 3-D representation of all contacts onto the AIM neuron morphology in an L1 animal, colored based on contacting partner identity, as labeled (right) in the detailed inset (black box) region. (D) AIM-PVQ contacts (in orange) and AIM-AVF contacts (in green), projected onto the AIM neurons (light purple) across developmental stages and augmented for clarity in the figure (see non-augmented contacts in Figure 8—figure supplement 1). Scale bar 2 μm.
Figure 5—figure supplement 1
Projecting contact profiles onto the segmented neuronal shapes.
(A–C) Graphical representations of the strategy utilized for creating the contact profiles for each of the adjacent neurons (purple, red, cyan) onto a cross section of the neuron of interest (Neuron A, yellow). (D–F) Electron micrograph from the L4 dataset with two adjacent neurons colored yellow and cyan. To build 3-D reconstructions of contact sites from adjacent neurons, we analyzed segmented neurons from the electron microscopy datasets in each slice (A, D). Each adjacent neuron is expanded in all directions to the pixel threshold distance (specified for each dataset; Supplementary file 1; Methods; CytoSHOW.org) (B, E). A new ROI (region of interest; purple, red, cyan in C; green in F) is created from the overlapping areas between the neuron of interest (yellow) and the adjacent neurons (C,F). (G–I) 3-D reconstruction of neuron (yellow) (G) with adjacent neuron (cyan), (H) with contact sites captured (green) across all slices, and (I) with contact areas from the adjacent neuron augmented (green) as seen in Figure 5D.
Figure 5—figure supplement 2
AIM contact sites.
Contact sites from PVQ (Orange and highlighted with orange arrowheads) and from AVF (Green and highlighted with green arrowheads) across developmental stages (as indicated) and projected onto the segmented AIM neurons (transparent purple). This figure is the unmodified NeuroSC outputs of contact profiles that correspond to Figure 5D. In Figure 5D these contact profiles were augmented. Scale bar = 2 um. See also Figure 5 and Figure 7—video 1.
Figure 6 with 2 supplements
Case study: Segmented morphologies of AIM, PVQ, and AVF across larval development.
(A) Cartoon schematic of the C. elegans head, pharynx (gray), and examined neurons with dashed black box representing the nerve ring region. (B) Schematic representation of the outgrowth path of the AVF neurons as observed by EM (Witvliet et al., 2021). The distal end of the AVFL neurite is highlighted with a green arrowhead in the schematic. (C) Neuronal morphologies of AIM (purple), PVQ (orange), AVFL (green) across postembryonic development, as indicated, with green arrowhead pointing to AVF outgrowth tip. Scale bar = 2 μm. Regions for insets (L1, dotted box; L2, dashed box) correspond to (D). Note that the AVF neuron class is comprised of a left and right counterpart. In C and D, we only show AVFL for simplicity. During development, AVFL and AVFR, as shown in A and B, both grow ipsilaterally along AIML in parallel and extend around the nerve ring together. This is unusual among classes of nerve ring neurons. (D) Morphologies of these neurons (rotated to the posterior view) display the AVFL neuron positions between the AIM and PVQ neurons at the L1 and L2 stage. Indicated outgrowth between neurons continues to the adult stage (Figure 6—video 1). Note how AVF outgrowth alters contact between PVQ and AIM (Figure 5D).
Figure 6—video 1
Analysis of AIM, PVQ, and AVF neuronal morphologies in developmental datasets.
3-D visualizations of AIM (Purple), PVQ (Orange), and AVF (Green) at (Left viewer) L1 (5 hours post hatching) and (Right viewer) L3 (27 hours post hatching) in NeuroSC. Note that at L1, AVF has not grown into the nerve ring, therefore, only AIM and PVQ are present, but by L3, the AVF neurons have grown between the AIM and PVQ neurons.
Figure 6—video 2
Navigating NeuroSC features that enable integration of Neurons, Contacts, and Synapses across developmental datasets.
Upon first opening NeuroSC, a tutorial will launch (Figure 8—figure supplement 4). In the NeuroSC menu, one can read about NeuroSC, access the tutorial, and navigate to the embryonic promoter database (Figure 8—figure supplement 4). The video shows the user searching neurons (AIM and PVQ) and adding neurons to the viewers (Figure 8—figure supplement 6). Side-by-side viewers with AIML, AIMR, and PVQL enable comparisons across developmental stages (L1, 0 hours post hatching and L4, 36 hours post hatching). Also shown in the video are the use of the in-viewer toolbar (Figure 8—figure supplement 7) and navigation menu (Figure 8—figure supplement 8) for object exploration.
Figure 7 with 2 supplements
Case study: AIM-PVQ and AIM-AVF synaptic positions across development.
(A) AIM-PVQ synaptic sites (dark orange arrowheads) and AIM-AVF synaptic sites (dark green arrowheads) in the segmented AIM neurons and reconstructed across postembryonic development from original connectomics data. Scale bar = 2 μm. (B) Schematic of the AIM, PVQ, and AVF circuitry across development based on synaptic connectivity and focusing on the stage before AVF outgrowth (L1), during AVF outgrowth (L2), and Adult; arrow direction indicates pre to post synaptic connection, and arrow thickness indicates relative number of synaptic sites (finest, <5 synapses; medium, 5–10 synapses; thickest, 11–30 synapses). (C) Zoom in of synaptic sites (green) in the Adult connectome and embedded into the AIM neuron morphology (light purple). In NeuroSC, presynaptic sites are displayed as blocks and postsynaptic sites as spheres, and a scaling factor is applied to the 3-D models (Materials and methods).
Figure 7—figure supplement 1
AVF synaptic sites.
Synaptic sites displayed onto transparent (green) AVF neurons across developmental stages. Presynaptic sites (spheres) and postsynaptic sites (blocks) are visualized between the AVF neurons and the AIM (purple) neurons, PVQ (orange) neurons and other AVF (either AVFL or AVFR; opaque green) neuron; Scale bar = 2 um.
Figure 7—video 1
Exploring contacts and synapses using NeuroSCAN.
Video of user navigating the tools of NeuroSC to examine synapses and contact profiles to yield results as in Figure 8—figure supplements 3 and 5. AIM neurons (Transparent Purple), AIM (Purple)-PVQ synaptic sites (Orange), and AIM-PVQ contact sites (Orange) at L1 (5 hours post hatching) are added into Viewer 1. AIM neurons (Transparent Purple), AIM(Purple)-PVQ synaptic sites (Orange), and AIM-PVQ contact sites (Orange), AVF (Green)-AIM synaptic sites, and AVF-AIM contact sites (Green) at L3 (27 hours post hatching) are added into Viewer 2. Contact sites and synaptic sites are compared across developmental stages by hiding AIM neurons. All contact sites for AIM are added for L1 (5 hours post hatching) into Viewer 3.
Figure 8 with 10 supplements
NeuroSC is a tool that enables integrated comparisons of neuronal relationships across development.
With NeuroSC, users have integrated access to: C-PHATE plots, 3-D morphological renderings, neuronal contact sites, and synaptic representations. Through stage-specific C-PHATE renderings, users can explore neuronal relationships from high dimensional contactome data. (Top) On C-PHATE plots, schematized here, each sphere represents an individual neuron, or a group of neurons clustered together during algorithm iterations. (Right) 3D renderings of select neurons can be visualized in the context of the entire nerve ring or other circuits (gray). (Left) AIM contact sites at L1 and the same region showing synapses. The inset shows a zoomed-in view of contacts and synapses - presynaptic sites (blocks) and postsynaptic sites (spheres). Data depicted here are from the L1 stage (0 hours post hatching).
Figure 8—figure supplement 1
Visualization of contact sites in NeuroSC.
(A) Search for a specific neuron (here, AIM) to filter (B) the list of contacts corresponding to the developmental slider. Neuron A (AIML, here) is the neuron onto which the contacts will be mapped. The Contacts drop-down menu sorts neurons alphabetically (here, colored according to the contact patch color in C). (C) 3-D reconstruction of all AIM contacts at L3 stage. See also Figure 6—video 2; Figure 7—video 1. In Figure 5D, contacts are augmented. Inset: Click on a contact rendering to show ’contact stats’. This pop-up displays quantifications of the selected contact relationship. Rank compares the summed surface area of contacts (‘patches’) between these two neurons relative to all other contact relationships for the primary neuron. A rank of 1 means this neuron pair shares the largest contact area. Total surface area is the total surface area in nm2 that covers the primary neuron in the contact relationship. Here the primary neuron is AIM. Contact area percentages are also shown for the cell and the whole nerve ring.
Figure 8—figure supplement 2
C-PHATE tutorial in NeuroSC.
(A) Add the C-PHATE plot corresponding to the position of the purple circle on the developmental slider (yellow box) by clicking (B) the + sign. (C) Screenshot of C-PHATE plot at L4 (36 hours post hatching), spheres represent individual neurons at the outer edge of the plot and DC iterations increase towards the center where spheres represent clusters of neurons and eventually the entire nerve ring. (D) Screenshot of C-PHATE plot at L4 (36 hours post hatching) with the spheres/clusters containing the AIM neurons highlighted (Blue) by selecting the AIM neurons within the light bulb menu (red box). See also Figure 4—video 1. NeuroSC features in this figure are not shown to scale.
Figure 8—figure supplement 3
Visualization of synaptic sites with NeuroSC.
(A) Search for synaptic sites for specific neuron(s) (e.g. AIM, PVQ) and choose a developmental time point with the slider. (B) Synapses dropdown menu contains a list of objects representing pre- and postsynaptic sites corresponding to all neuron names in the search bar and sorted alphabetically. Searched neurons can be used with the synaptic filter (C) to select for synapse type (electrical or chemical; Note: only use this feature for L4_36 hours post hatching and Adult_48 hours post hatching) and to filter objects by synaptic specialization (pre or post; gray dotted box), (D) which will follow the filter logic (example shown for AIM and PVQ). (E) To enable visualization of subsets of synapses and differentiate between pre- and postsynaptic sites, each synapse contains object(s) representing the postsynaptic site(s) as spheres (Blue and Purple) and the presynaptic site as a block (Orange). These are ordered ‘by synapse’, with all postsynaptic objects, then the presynaptic object. This specific example corresponds to a 3-D representation of the PVQL (Orange, Pre) AIAL (Blue, Post), AIML (Purple, Post) synapse. (F) All synaptic sites contain the name of the presynaptic neuron (Orange), neuron type (chemical, electrical, or undefined), list of postsynaptic neuron(s) (Blue), and Unique identifier (Black; Section, letter) for cases with multiple synapses between the same neurons. The ‘section’ is unique to each synapse between specified neurons and at that specific developmental stage. It is listed in order of its anteroposterior position in the neuron. Synapse names are not linked through developmental datasets. If the synapse is polyadic, there will be multiple postsynaptic neuron names and objects associated with a single presynaptic site. See also Figure 7—video 1. (G) Right-click on a rendered synapse to open synapse stats. Expand (click arrow in green box) to show the numbers of synapses, broken down by polyadic variations if applicable, per the primary neuron (in this case the primary neuron is PVQL).
Figure 8—figure supplement 4
Opening page view and menu.
(A) View of opening page. (B) Menu for access to the ‘About’ window for referencing source information, the Tutorial, and the developmental Promoter database. See also Figure 6—video 2.
Figure 8—figure supplement 5
The NeuroSC interface enables interrogation of neuronal relationships across development.
(A) The left-facing arrow to minimize the left panel and optimize space for the viewer windows. The interface contains four main parts: (B–E) Filters, (F–J) Results, (K–M) Viewer Navigation, and (N–Q) viewer windows. Filter Results by (C) searching for neuron names, (D) selecting a dataset with the developmental slider (in hours post- hatching), (E) and filtering synapses based on the pre- or post-synaptic partner on the neurons that are on the search bar. (F) Results drop down menus (filtered by B) for (G) Neuronal morphologies (shown in the viewer as purple in (O)), (H) Contacts (shown in green (O)); (I) Synapses (shown in Orange in (O)); and (J) C-PHATE (shown in (Q)), which gets filtered by the developmental slider in (D). (K) Viewer Navigation to rotate the 3-D projections in all viewers simultaneously (Play All) and which contains a drop-down menu for each viewer (L,M). The viewers are named as Viewer 1 (L, N) or CPHATE viewer (M, P) and followed by information of the developmental stage and the hours post hatching for the objects in the viewer. (O) Reconstruction of the AIM neurons with AVF contacts and synapses at L3 27 hours post hatching; scale bar = 2 µm. (Q) C-PHATE plot at L1 (0 hours post hatching). See also Figure 6—video 2.
Figure 8—figure supplement 6
Select and Add objects to viewers.
(A) Click ‘select (number) items’ to select all items in the drop down list (green box), or (A’) click the hexagon next to each item (green box). (B) Click ‘Add Selected’ (purple box) to add all selected items or (B’) click ‘Add to’ (purple box) to add each item individually. (C) To add the selected item(s) to an existing viewer of the same developmental stage or to a new viewer, choose a viewer as indicated. (D) Click ‘Deselect (number) items’ (orange box) to deselect items. See also Figure 6—video 2.
Figure 8—figure supplement 7
In-viewer toolbar features.
(A) In-viewer toolbar for Neurons, Contacts, and Synapses and C-PHATE (shown here, only Neurons). (B, K) Change the background color of viewer from dark (white box, moon) to white (white box, sun). (C, L) Change the color of any objects by selecting a desired color, transparency, or color code and selecting the object (or instance) name (here, AIML and AIMR). (D, M) Copy and paste your neuron and contact profile scenes through time into a comparative side by side window changing the developmental stage for items in the viewer by using the in-viewer developmental slider. Note a new window will be generated, which can be dragged next to the original window.(N) Add 3-D representations of the Nerve Ring for that developmental stage, note: must be added to the scene last. (E, O) Record and download movies for the viewer. (F,P) Download.gltf files and viewer screenshot (png). (G) Rotate objects around the y-axis. (H) Zoom in and (I) zoom out, and (J) reset objects to original positions in the viewer. See also Figure 6—video 2.
Figure 8—figure supplement 8
Viewer navigation menu.
(A) Navigation bar contains a drop-down menu for each viewer (shown here, six viewers at varied developmental stages) and a ‘Play all’ button for simultaneously rotating all objects in each viewer around the y-axis (Figure 6—video 2). Each viewer drop-down menu contains a drop-down menu for Neurons (green box), Contacts, and Synapses. (B) Viewer 6 with reconstructions of three neurons (AIML and AIMR, purple; PVQL, orange) at Larval Stage 4 (L4), 36 hours post hatching. (C) Browse and Select objects in the viewer by navigating the nested drop-down menus. (D) Manage objects in viewers with options to select, group, hide, and delete objects in each viewer. Objects can be deleted with ‘select’ and keyboard ‘delete’. See also Figure 6—video 1.
Figure 8—figure supplement 9
NeuroSC architecture.
(A) Source data are defined in a file tree structure that contains various assets such as .gltf files representing various entities, as well as CSVs storing entity metadata. The directory structure outlines a vertical hierarchy starting at the developmental stages, then branching downwards through neuron, C-PHATE, contact, and synapse data.(B) A Golang script can be invoked to traverse the directory tree and ingest the files. This parses the hierarchical file path and file contents, verifying the data and associating it with underlying entities, writing it to the database.(C) The backend consists of a Postgres Database to store underlying data, a Block Storage Volume that houses static assets (.gltf files, javascript bundle, css styles, etc), and a Golang application. The Golang application surfaces a JSON REST API, leveraging caching strategies to reduce DB load. It also serves the static assets to the client from the block storage. A variable number of Virtual Machines run the backend application code, scaling as needed to accommodate traffic. (D) The client-side is a React application that uses Geppetto for entity rendering (Cantarelli et al., 2018). User interactions on the frontend result in queries to the backend to filter, sort, and search via the REST API, resulting in the entities and metadata to be rendered to an interactive canvas.
Figure 8—figure supplement 10
NeuroSC data model.
(A) Reference scheme for B-F; Instance refers to the category (e.g., B, Neuron; C, Developmental Stage), which contains a name or identifier (id) for each object, lists of files associated with the instance (C, Developmental Stage does not have files), and metadata to further describe each instance, which is usually a string (str) or an integer (int). (B) The neuron name is the foundation for the Contacts, Synapses, and C-PHATE, which enables integration across each of these representations and across developmental stages (time points) with metadata from WormAtlas (wormatlas.org/MoW_built0.92/MoW.html). (C) The Developmental Stages are named by the larval stages (L1, L2, L3, L4, Adult), and the metadata captures the list of time points within those developmental stages (i.e. L1, 0 hours post hatching, and L1, 5 hours post hatching). (D) C-PHATE objects are named with a list of Neurons. (E) Contacts link to the Neuron names (Neuron A and Neuron B nomenclature in Figure 8—figure supplement 1), and metadata annotates the surface area in nm2 of contact quantified in the source Electron Microscopy micrographs. (F) Synapses link to the Neuron names (Pre, type (chemical, electrical or undefined), Post, and section described in Figure 8—figure supplement 3).
Additional files
-
Supplementary file 1
Nerve ring regions, resolutions, and pixel threshold distances used to calculate adjacency matrices and to create contact sites for each dataset.
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp1-v1.xlsx
-
Supplementary file 2
Scaling factors and rotation corrections for 3-D representations of Neurons, Contacts, and Synapses for each dataset.
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp2-v1.xlsx
-
Supplementary file 3
Stratum 1 (Red) Sankey diagrams of clustered neurons for each Diffusion Condensation iteration in each dataset.
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp3-v1.xlsx
-
Supplementary file 4
Stratum 2 (Purple) Sankey diagrams of clustered neurons for each Diffusion Condensation iteration in each dataset.
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp4-v1.xlsx
-
Supplementary file 5
Stratum 3 (Blue) Sankey diagrams of clustered neurons for each Diffusion Condensation iteration in each dataset.
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp5-v1.xlsx
-
Supplementary file 6
Stratum 4 (Green) Sankey diagrams of clustered neurons for each Diffusion Condensation iteration in each dataset.
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp6-v1.xlsx
-
Supplementary file 7
Sankey diagrams of AIM, PVQ, and AVF containing clusters for each Diffusion Condensation iteration in each dataset.
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp7-v1.xlsx
-
Supplementary file 8
L1 (0 hours post hatching) adjacency counts and searchable counter for summed adjacencies.
Type the name of a ‘Neuron of Interest’ (NOI) in the indicated cell to filter for the summed adjacency counts for each contact partner. For each partner, there are two columns: Total number of contacts (number of EM sections NOI and partner are in contact) and Total Weights (summed number of pixels NOI and partner contacts).
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp8-v1.xlsx
-
Supplementary file 9
L1 (5 hours post hatching) adjacency counts and searchable counter for summed adjacencies.
Type the name of a ‘Neuron of Interest’ (NOI) in the indicated cell to filter for the summed adjacency counts for each contact partner. For each partner, there are two columns: Total number of contacts (number of EM sections NOI and partner are in contact) and Total Weights (summed number of pixels NOI and partner contacts).
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp9-v1.xlsx
-
Supplementary file 10
L2 (23 hours post hatching) adjacency counts and searchable counter for summed adjacencies.
Type the name of a ‘Neuron of Interest’ (NOI) in the indicated cell to filter for the summed adjacency counts for each contact partner. For each partner, there are two columns: Total number of contacts (number of EM sections NOI and partner are in contact) and Total Weights (summed number of pixels NOI and partner contacts).
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp10-v1.xlsx
-
Supplementary file 11
L3 (27 hours post hatching) adjacency counts and searchable counter for summed adjacencies.
Type the name of a ‘Neuron of Interest’ (NOI) in the indicated cell to filter for the summed adjacency counts for each contact partner. For each partner, there are two columns: Total number of contacts (number of EM sections NOI and partner are in contact) and Total Weights (summed number of pixels NOI and partner contacts).
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp11-v1.xlsx
-
Supplementary file 12
L4 (36 hours post hatching) adjacency counts and searchable counter for summed adjacencies.
Type the name of a ‘Neuron of Interest’ (NOI) in the indicated cell to filter for the summed adjacency counts for each contact partner. For each partner, there are two columns: Total number of contacts (number of EM sections NOI and partner are in contact) and Total Weights (summed number of pixels NOI and partner contacts).
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp12-v1.xlsx
-
Supplementary file 13
Adult (48 hours post hatching) adjacency counts and searchable counter for summed adjacencies.
Type the name of a ‘Neuron of Interest’ (NOI) in the indicated cell to filter for the summed adjacency counts for each contact partner. For each partner, there are two columns: Total number of contacts (number of EM sections NOI and partner are in contact) and Total Weights (summed number of pixels NOI and partner contacts).
- https://cdn.elifesciences.org/articles/103977/elife-103977-supp13-v1.xlsx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/103977/elife-103977-mdarchecklist1-v1.pdf
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Exploring neurodevelopment via spatiotemporal collation of anatomical networks with NeuroSC
eLife 13:RP103977.
https://doi.org/10.7554/eLife.103977.3
