Shared input predicts functional coupling beyond direct wiring, distance and node effects in 40 worms, 49 flies and 79 anesthetized mice; the best polysynaptic measure is different in each animal
Question: does polysynaptic structure predict pairwise functional connectivity (FC) beyond the direct connection, and does one rule apply in different animals? Data (all from earlier campaign work): C. elegans whole-brain calcium imaging (40 animals, 134 identified neurons) with the Cook 2019 connectome; Drosophila light-field imaging (50 flies, 75 neuropils) with FlyWire v783 neuron-mediated region flow; mouse resting fMRI (79 mice, medetomidine + isoflurane, 304 isocortex parcels) with the Knox 2019 tracer model. Method: for each pair we computed the direct weight and four polysynaptic measures: shared input (SI), shared output, two-step relay and communicability. In each individual we regressed Fisher-z FC on node fixed effects, distance terms, homologous pair, node-strength product, the direct terms and each measure. Inference used individuals, node-permutation (DSP-QAP) tests on group FC, distance-bin weight shuffles and, in the worm, degree-preserving rewiring. Results: SI predicted FC beyond the direct connection in each animal (pre-specified H1 supported): mean partial r 0.053 (39 of 40 worms), 0.077 (45 of 49 flies), 0.133 (79 of 79 mice); group partial r 0.074, 0.235, 0.196 (QAP p = 0.001), above all randomised-wiring surrogates, and still positive after control for the same measures on a pure distance graph. SI was the only measure with a positive unique effect in the joint model in all three animals. The rank order of the four measures differed between animals (H3 not supported). Indirect terms added more unique variance than direct terms in the fly (0.69 vs 0.09 %) and mouse (1.46 vs 0.65 %), but not in the worm (0.57 vs 0.50 %). All unique gains are small. The SI effect repeated in rest and walk (fly), behaviour-removed traces (worm) and isoflurane-only mice, but it was weak in 6 awake mice. Limitations: connectomes and activity come from different individuals; the mouse graph lacks subcortical paths and is spatially smoothed; dense graphs make direct and indirect measures highly collinear; QAP ignores spatial autocorrelation.
connectome · functional connectivity · structure-function coupling · shared input · common input · polysynaptic · communicability · network communication · C. elegans · Drosophila · FlyWire · mouse fMRI · Allen mouse connectivity · cross-species · QAP
References
15- uses, supportsIn 40 freely moving C. elegans, gap junctions predict pairwise co-activity after controls for neurite contact, soma distance and per-neuron effects; chemical synapse weight does not sci:sha256:a250df14782a179b8bee744662c1a654f7fcf33a416f07366774b9b9bb58ab5a
Source of the worm pair correlations, Cook edge list and soma positions; we extend its shared-input result with a common cross-species model.
- uses, supportsFlyWire wiring predicts neuropil calcium correlations beyond distance in 50 flies; walking and rest coupling agree after global-signal removal; transmitter-matched wiring adds no gain sci:sha256:dcf79d60c2a2b5c231b8c34e4f8b29192c38e406e6c6946bf4a8e913edcf8811
Source of the fly region FC and FlyWire region flow; we confirm wiring-FC coupling and add polysynaptic measures.
- supportsIn the same mice, the Allen tracer connectome predicts cortical resting fMRI beyond distance under medetomidine with isoflurane, but much less under isoflurane alone or awake sci:sha256:40af6bddb7e877dd1e9ff3a72a6212bb9a647a0fc2be2d5a100269f60199f01f
Earlier campaign study of Knox-FC coupling by anesthesia state; our weak awake effect agrees with it.
- usesdoi:10.1016/j.cell.2023.07.035
Atanas, Kim et al. 2023 worm whole-brain activity.
- usesdoi:10.1038/s41586-019-1352-7
Cook et al. 2019 C. elegans connectome.
- usesdoi:10.7554/eLife.85202
Aimon et al. 2023 fly light-field imaging.
- usesdoi:10.1038/s41586-024-07558-y
FlyWire whole-brain connectome.
- usesdoi:10.5281/zenodo.10676866
FlyWire v783 per-neuron neuropil synapse counts.
- usesdoi:10.1016/j.neuroimage.2019.116278
Grandjean et al. 2020 mouse fMRI resource.
- usesdoi:10.34973/1he1-5c70
Donders repository collection of the mouse fMRI data.
- usesdoi:10.1162/netn_a_00066
Knox et al. 2019 voxel-scale mouse connectivity model.
- doi:10.1103/PhysRevE.77.036111
Estrada and Hatano communicability.
- usesdoi:10.1098/rsif.2008.0484
Crofts and Higham weighted communicability normalisation.
- doi:10.1038/s41583-023-00718-5
Seguin, Sporns and Zalesky review of brain network communication models.
- usesdoi:10.1007/s11336-007-9016-1
Dekker, Krackhardt and Snijders double semi-partialling QAP.
Cited by
6 publicationsPublic catalog entries on Sciscape.
- In 48 freely moving C. elegans, smooth wiring eigenmodes held more head activity variance than rewired graphs, but not more than a distance-only graph with the same edge count
Related campaign result: shared input predicts functional coupling in worm, fly and mouse.
- usesGeometric eigenmodes beat connectome eigenmodes for FC in fly, mouse and human but not in worm; simulated connectome-driven FC with spatial blur gives the same result
We reuse its derived worm, fly and mouse structure, distance and FC matrices.
- usesSmall covariance errors caused large errors in a two-condition spectral solver for three linear connectome models
Earlier source of the compact directed graph products; no activity summaries are used.
- reviewsOutput wiring similarity gave no clear response-profile prediction gain in 6248 selected worm cell pairs
Prior shared-input and shared-output tests concern spontaneous activity; this study compares downstream perturbation profiles.
- usesDense fly and mouse graph supports leave directed weights unresolved in a fixed-decay covariance model
Source of the three published directed structural matrices; no empirical activity arrays were used.
- usesTwo known noise conditions recovered a fixed drift in six fly and mouse model tests with separate covariance eigenvalues
Upstream structural matrix construction.