Neural Networks of the Mouse Neocortex

Neural Networks of the Mouse Neocortex Zingg Hintiryan Gou Song Bay Bienkowski Foster Yamashita

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Neural Networks of the Mouse Neocortex 2014 Cell 10.1016/j.cell.2014.02.023 Zingg2014-neural-networks-of-the-mouse-neocortex Zingg2014-neural-networks-of-the-mouse-neocortex.txt, the project's staged copy 7f18f8a318fe757ea1f49579e1ae9e8ea5747f1d not-redistributable neuroscience

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them how areas across the entire mouse neocortex organise into communication networks, mapped by over 600 traced pathways Quotation withheld because this record is not cleared for redistribution. how do different cortical areas communicate at the network level across the whole structure Quotation withheld because this record is not cleared for redistribution. delivers a cortical connectivity atlas and matrices from uniformly collected tracer data, and finds the neocortex clusters into four somatic sensorimotor, two medial and two lateral subnetworks Quotation withheld because this record is not cleared for redistribution. existence Quotation withheld because this record is not cleared for redistribution. cortico-cortical connectivity of the mouse neocortex Quotation withheld because this record is not cleared for redistribution. system-or-tool-building system-or-tool-building Quotation withheld because this record is not cleared for redistribution. each injection is designed to return four kinds of information at once - inputs, outputs, reciprocal connections and intermediate stations - by pairing an anterograde with a retrograde tracer Quotation withheld because this record is not cleared for redistribution. the two tracer directions are used as independent measurements of the same connection, and their agreement is reported as a validity check Quotation withheld because this record is not cleared for redistribution. injection specificity is cross-validated by tracing the reverse direction: retrograde tracer into the regions the anterograde labelling had targeted Quotation withheld because this record is not cleared for redistribution. the subnetworks are not stipulated but produced by reordering the matrix with a clustering algorithm, and named afterwards Quotation withheld because this record is not cleared for redistribution. the clustering makes the matrix diagonal readable as within-subnetwork density and the off-diagonal as between-subnetwork traffic Quotation withheld because this record is not cleared for redistribution. the whole analysis rests on hand annotation, because no tool can identify fine anatomic boundaries, and the strengths recorded are semi-quantitative Quotation withheld because this record is not cleared for redistribution. each pathway is rendered as a separate layer in a common anatomic frame so that arbitrary combinations can be viewed together Quotation withheld because this record is not cleared for redistribution. uniform collection and analysis is claimed as the advantage over connectivity matrices assembled from published literature Quotation withheld because this record is not cleared for redistribution. the somatosensory subdivisions used as network nodes are themselves validated against known somatotopy in the brainstem before being used Quotation withheld because this record is not cleared for redistribution. a visualisation artefact is disclosed rather than left for the reader to discover: the injection sites look larger than they are once the image is brightened enough to show fine fibers Quotation withheld because this record is not cleared for redistribution. the resource is framed as the input to later causal work, naming the specific techniques that would test any mapped pathway Quotation withheld because this record is not cleared for redistribution. Interpreting large-scale connectivity data remains an unsolved challenge. what this limits: the interpretive claims, not the resource: the paper demonstrates THE FEASIBILITY OF PRODUCING AND COLLECTING LARGE-SCALE CONNECTIVITY DATA, and names interpretation as the unsolved part Quotation withheld because this record is not cleared for redistribution. Published injection-site images overstate the true injection size. what this limits: any reading of injection extent from the published figures: WHEN IMAGES ARE ADJUSTED TO REVEAL FINE FIBERS, INJECTION [extent appears larger] than the ~250-500um stated Quotation withheld because this record is not cleared for redistribution. Annotation had to be done by hand because no automated tool can delineate fine anatomical boundaries. what this limits: the precision of every areal boundary in the atlas: INFORMATICS TOOLS THAT AUTOMATICALLY AND PRECISELY IDENTIFY FINE ANATOMIC BOUNDARIES ... ARE NON-EXISTENT, so annotation is by hand Quotation withheld because this record is not cleared for redistribution. The mapped long-range interactions have not been validated at the synaptic or cell-type level. what this limits: the synaptic and cell-type reading of any mapped connection: these MAY BE VALIDATED AT THE SYNAPTIC LEVEL USING TRANSSYNAPTIC VIRAL TRACING - not done here Quotation withheld because this record is not cleared for redistribution. interpreting connectivity data at this scale is named as the unsolved part, not the collection of it what this limits: the same scope as limitations[0] - collection is demonstrated, interpretation is not Quotation withheld because this record is not cleared for redistribution. no automated tool exists for the annotation step, so the atlas is hand-made and semi-quantitative what this limits: the same scope as limitations[2], with the consequence: the analysis is hand-made and semi-quantitative Quotation withheld because this record is not cleared for redistribution. the matrices discard route, laminar specificity and topography, which the authors call critical features of networks what this limits: every use of the connectivity matrices: THEY EXCLUDE DETAILS LIKE PROJECTIONS ROUTES, LAMINAR SPECIFICITY OF PROJECTIONS, OR TOPOGRAPHICAL organisation - the condensed view drops features the authors call critical Quotation withheld because this record is not cleared for redistribution. the mapped connections are anatomical and require separate validation at the synaptic and functional level what this limits: the same scope as limitations[3] - anatomical connections, awaiting synaptic and functional validation Quotation withheld because this record is not cleared for redistribution. human network findings are held to need validation in animal models at higher resolution, which is the gap this atlas is built to fill what this limits: the transfer of human network findings: IT IS NECESSARY TO VALIDATE THESE NETWORKS USING RELIABLE NEURAL TRACT TRACING METHODS IN ANIMAL MODELS AT A HIGHER RESOLUTION Quotation withheld because this record is not cleared for redistribution. the hub - a region of high network interaction - imported to interpret the claustrum and lateral entorhinal area Quotation withheld because this record is not cleared for redistribution. reordering nodes so that dense groups fall along the diagonal, which converts a connectivity table into a set of named subnetworks Quotation withheld because this record is not cleared for redistribution. cross-validating anterograde injection specificity by applying retrograde tracers to the regions targeted by anterogradely labelled axon terminals, and the reverse eliminates: that the anterograde labelling reflects nonspecific injection spread rather than a genuine projection · outcome: held Quotation withheld because this record is not cleared for redistribution. connectome - as the C. elegans-style complete synaptic wiring diagram versus as the meso-scale regional map this paper builds Quotation withheld because this record is not cleared for redistribution. cortical network - as derived from human fMRI and diffusion imaging versus as traced anatomically in an animal model; the paper treats the second as validation of the first Quotation withheld because this record is not cleared for redistribution. connectivity matrix - assembled piecemeal from published literature versus collected and analysed uniformly; the paper claims the difference matters Quotation withheld because this record is not cleared for redistribution. subnetwork - as a cluster in a reordered matrix versus as an anatomically or functionally defined system; here the first produces the second, and the names come last Quotation withheld because this record is not cleared for redistribution. primary sensory cortex as strictly unisensory versus as cross-modally modulated - a boundary the connectivity data is said to challenge Quotation withheld because this record is not cleared for redistribution. connection strength - as a semi-quantitative hand-graded density, which is what the weights in every matrix here actually are Quotation withheld because this record is not cleared for redistribution. claim_type: method a second reading DIFFERING from the recorded value. Recorded: existence — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. method: double co-injection tract tracing, one anterograde and one retrograde tracer per site, across the whole neocortex, hand-annotated into weighted directional matrices and clustered a second reading DIFFERING from the recorded value. Recorded: system-or-tool-building — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. phenomenon: large-scale cortico-cortical network organisation a second reading DIFFERING from the recorded value. Recorded: cortico-cortical connectivity of the mouse neocortex — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. builds-on → the double co-injection tract tracing method of Thompson and Swanson (2010) Quotation withheld because this record is not cleared for redistribution. extends → cortical connectivity matrices previously assembled from literature data (Honey et al., 2007; Sporns et al., 2007; Markov et al., 2010) Quotation withheld because this record is not cleared for redistribution. builds-on → the concept of cross-modal modulation (Driver and Noesselt, 2008; Stein and Stanford, 2008) Quotation withheld because this record is not cleared for redistribution. extends → the account of the lateral entorhinal cortex as a gateway to the hippocampus (de Curtis and Paré, 2004) Quotation withheld because this record is not cleared for redistribution. builds-on → the experimental procedures of Hintiryan et al. (2012) Quotation withheld because this record is not cleared for redistribution. critiques → cortical connectivity matrices assembled piecemeal from the published literature differs in: THE AXIS IS HOW THE DATA WERE ASSEMBLED. 'the networks reported here are based on data collected and analyzed in a homogenous fashion rather than gathered piecemeal from the literature.' A critique of provenance rather than of any result - and one this project should read against itself, since a cross-field corpus is assembled exactly the way this sentence criticises. Quotation withheld because this record is not cleared for redistribution. extends → human structural and functional network studies using fMRI and diffusion tensor imaging, which the paper holds need animal validation at higher resolution Quotation withheld because this record is not cleared for redistribution. contradicts → the assumption that mammalian primary sensory cortices are strictly unisensory Quotation withheld because this record is not cleared for redistribution. builds-on → the network-science framing of brain organisation, including the hub concept Quotation withheld because this record is not cleared for redistribution. extends → wiring diagrams previously assembled only for specific functional systems or smaller scales Quotation withheld because this record is not cleared for redistribution. subnetworks —operationalised-as→ clustering algorithm subnetworks clustering algorithm strength of the relation: 0.95 Audit 2026-09-28: DOUBTFUL - read against its span and kept, but subnetworks were revealed by a clustering algorithm; operationalised-as is loose. Quotation withheld because this record is not cleared for redistribution. claustrum —is-a→ hubs claustrum hubs strength of the relation: 0.8 Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span notes the claustrum's broad intracortical connections; is-a 'hubs' is inferred. Quotation withheld because this record is not cleared for redistribution. medial —distinct-from→ lateral medial lateral strength of the relation: 0.9 Quotation withheld because this record is not cleared for redistribution. lateral entorhinal area —part-of→ trisynaptic circuit lateral entorhinal area trisynaptic circuit strength of the relation: 0.85 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span writes 'ENTl'; the endpoint is a variant, and part-of is loose ('the starting point of' the circuit). Quotation withheld because this record is not cleared for redistribution. cluster analysis cluster analysis Quotation withheld because this record is not cleared for redistribution. percentage agreement between two methods percentage agreement between two methods Quotation withheld because this record is not cleared for redistribution. weighted directional connectivity matrices weighted directional connectivity matrices Quotation withheld because this record is not cleared for redistribution. clustering algorithm clustering algorithm Quotation withheld because this record is not cleared for redistribution. semi-quantitative densities semi-quantitative densities Quotation withheld because this record is not cleared for redistribution. The mouse neocortex resolves into eight subnetworks: four somatic sensorimotor, two medial and two lateral, each with its own topology, communicating through selected cortical areas. Quotation withheld because this record is not cleared for redistribution. Anterograde and retrograde tracing agreed on 89% of matrix entries, cross-validating the connectivity dataset. Quotation withheld because this record is not cleared for redistribution. The claustrum's cortical connectivity is asymmetric: it receives bilateral cortical input but sends output almost exclusively to the same hemisphere. Quotation withheld because this record is not cleared for redistribution. The temporal association, ectorhinal and perirhinal areas are back-labelled from layers 2 and 5a of the entire neocortex except ACAv, RSP and VISp. Quotation withheld because this record is not cleared for redistribution. Because every main node of the four somatic sensorimotor subnetworks is reciprocally connected, sensory and motor areas can interact directly without routing through higher-order association cortex. Quotation withheld because this record is not cleared for redistribution. The findings support cross-modal modulation, challenging the idea that mammalian primary sensory cortices are strictly unisensory. Quotation withheld because this record is not cleared for redistribution. The second medial subnetwork is a serial chain running from dorsal subiculum to RSPv to ACAv and then to the infralimbic, prelimbic and medial orbitofrontal areas. Quotation withheld because this record is not cleared for redistribution. Ventral retrosplenial cortex is the sole neocortical area receiving dense dorsal subicular input, making it the entry point for dorsal hippocampal information into neocortex. Quotation withheld because this record is not cleared for redistribution. The dorsolateral prefrontal cortex is where all four somatic sensorimotor subnetworks converge. Quotation withheld because this record is not cleared for redistribution. The ventromedial prefrontal cortex is where the medial and lateral subnetworks converge, letting external sensory and internal visceral/gustatory information be integrated. Quotation withheld because this record is not cleared for redistribution. 89% of connections were labelled by both the anterograde and retrograde methods, so the two datasets largely agree Quotation withheld because this record is not cleared for redistribution. clustering placed every node into a small number of distinct cortico-cortical modules Quotation withheld because this record is not cleared for redistribution. the four somatic sensorimotor modules each formed a separate subnetwork Quotation withheld because this record is not cleared for redistribution. the claustrum and lateral entorhinal area have broad unique connections consistent with a hub role Quotation withheld because this record is not cleared for redistribution. claustral connectivity is asymmetric: inputs bilateral, outputs almost exclusively ipsilateral Quotation withheld because this record is not cleared for redistribution. the temporal subnetwork receives input from nearly the entire neocortex Quotation withheld because this record is not cleared for redistribution. within the somatic sensorimotor subnetworks, sensory and motor areas interact directly without higher-order association areas Quotation withheld because this record is not cleared for redistribution. the lateral subnetworks are points of massive convergence, each connected to a large number of cortical areas Quotation withheld because this record is not cleared for redistribution. each subnetwork converges onto a discrete territory of prefrontal cortex Quotation withheld because this record is not cleared for redistribution. the prefrontal zones are relatively segregated, which may help define their unique processing roles Quotation withheld because this record is not cleared for redistribution. about 600 pathways from around 317 co-injections are available, 304 efferent and 296 afferent Quotation withheld because this record is not cleared for redistribution. the findings support cross-modal modulation, against strictly unisensory primary sensory cortices Quotation withheld because this record is not cleared for redistribution. macro- or meso-connectome macro- or meso-connectome Quotation withheld because this record is not cleared for redistribution. hub hub Quotation withheld because this record is not cleared for redistribution. iConnectome iConnectome Quotation withheld because this record is not cleared for redistribution. meso-connectome meso-connectome Quotation withheld because this record is not cleared for redistribution. double co-injection double co-injection Quotation withheld because this record is not cleared for redistribution. hubs hubs Quotation withheld because this record is not cleared for redistribution. clustering clustering Quotation withheld because this record is not cleared for redistribution. intraconnectivity intraconnectivity Quotation withheld because this record is not cleared for redistribution. interconnectivity interconnectivity Quotation withheld because this record is not cleared for redistribution. injection site injection site Quotation withheld because this record is not cleared for redistribution. density of intraconnectivity density of intraconnectivity Quotation withheld because this record is not cleared for redistribution. laminar distribution patterns laminar distribution patterns Quotation withheld because this record is not cleared for redistribution. anterograde and retrograde labeling anterograde and retrograde labeling Quotation withheld because this record is not cleared for redistribution. Over 600 — labelled neuronal pathways collected across the whole mouse neocortex Quotation withheld because this record is not cleared for redistribution. 240 — intracortical connections manually reconstructed into the common atlas frame Quotation withheld because this record is not cleared for redistribution. 89% — share of matrix entries labelled by both the anterograde and the retrograde method Quotation withheld because this record is not cleared for redistribution. ~600 pathways (304 efferent and 296 afferent), ~317 co-injections — contents of the released dataset at time of publication Quotation withheld because this record is not cleared for redistribution. 80 anterograde and 160 retrograde — pathways included in the reconstructed cortico-cortical connectivity map Quotation withheld because this record is not cleared for redistribution. ~ 250-500μm — diameter of the tracer co-injection sites Quotation withheld because this record is not cleared for redistribution. 50μm — section thickness used for all brains Quotation withheld because this record is not cleared for redistribution. over 600 labeled neuronal pathways — the atlas contents Quotation withheld because this record is not cleared for redistribution. 240 — intracortical connections manually reconstructed into the common frame Quotation withheld because this record is not cleared for redistribution. 89% — connections labelled by both tracer directions Quotation withheld because this record is not cleared for redistribution. ~250-500μm — co-injection size, mostly confined within individual cortical areas Quotation withheld because this record is not cleared for redistribution. 304 efferent and 296 afferent — pathway counts from about 317 co-injections Quotation withheld because this record is not cleared for redistribution. 8 week old male C57Bl/6J mice — the animals Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.1016/j.cell.2014.02.023.