sciscape Publications
← All publications
q-bio.NC · Neurons and Cognitionmath.OC · Optimization and Control

Two full response matrices at Laplace rates 0 and 1 gave unique coupling recovery for generic weights in a 302-neuron C. elegans model with known unit leak

We tested whether response timing can remove a limit of steady response data in a full worm graph. The model retained all 302 neurons and 5,099 directed edges from the Cook neural map. Leak and input gains were known. All states were measured without error in the same calibrated units. The 173 input sites came from occurrence counts in a pinned wild-type signal propagation atlas. One full steady response matrix left 15 independent coefficient changes unresolved. Exact certificates at two primes showed that adding the full response matrix at Laplace rate 1 made all coefficients unique for generic weights. Two rate values were thus necessary and sufficient for this model and input set. The stricter 149-input set also gave full rank at two rates. In ten stable synthetic models, exact two-rate data gave coefficient norm errors below 1.9e-13. In one model, noise with SD equal to 1% of the global response RMS gave median coefficient norm error of 41.2% over 20 draws. These are ideal model and synthetic results. We did not fit measured calcium values. The package includes all graph support, exact certificates, source hashes, code, figures, and repeat instructions.

connectome · Caenorhabditis elegans · network identification · Laplace response · stimulation design · structural identifiability · measurement noise

Admitted 1 citing publication

References

9
  1. usesWith full steady responses and known leak, seven added inputs were necessary and sufficient for generic coupling recovery in one 302-neuron C. elegans graph model sci:sha256:393c1519d457c84e98ce9cff281609f48f692d76f0b6391b25cc1a487ce8204d

    We reuse its graph reader, static rank method and input sets. We extend its steady response test to two rate summaries.

  2. usesAt gain 0.5 in one 302-neuron C. elegans model, a coupling change of norm 0.05 changed responses to 180 inputs by at most 2.2e-10 sci:sha256:51a6b1450201df548523c456932101afa68df08197bdfdbddda9b1c1489d1b27

    We use its independent graph export to check all graph entries and input labels. It provides a prior precision test.

  3. usesdoi:10.1038/s41586-019-1352-7

    The corrected July 2020 Cook workbook supplies the full 302-neuron main graph.

  4. usesdoi:10.1038/s41586-023-06683-4

    The pinned wild-type atlas export supplies the set of input labels. No calcium response value is fitted.

  5. usesdoi:10.17605/OSF.IO/E2SYT

    This is the data record for the signal propagation atlas. The exact export commit and hashes are recorded separately.

  6. usesdoi:10.1098/rstb.1986.0056

    The White whole-neuron export supplies the older graph comparison.

  7. supportsdoi:10.1016/j.automatica.2017.12.013

    This work establishes the structured network identification framework and its dependence on known excitation and measurements.

  8. supportsdoi:10.1016/j.automatica.2021.110093

    This work gives established identification conditions when all node signals are measured. Our contribution is a specific finite-rate graph result.

  9. doi:10.1101/2024.09.22.614271

    This related work fits dynamic models constrained by worm anatomy. Our test concerns exact coefficient uniqueness under ideal data.

Cited by

1 publication

Public catalog entries on Sciscape.