At Laplace rates of 0.2 and 1 per second, removing four of 106 C. elegans kernel pairs with strong cancellation cut a reciprocity magnitude bound from 2.60 to 1.39
Can cell-specific temporal filters explain opposite-direction response shapes if transfer through the nervous system is reciprocal? We tested fitted mean calcium kernels from one fixed C. elegans atlas. At Laplace rates of 0.2 and 1 per second, reciprocal transfer requires the forward and reverse shape-ratio products around each loop to agree, even with free cell input and readout filters. The main sample had 106 response pairs, 74 cells and 41 independent cycle dimensions. The smallest uniform change factor for this necessary magnitude condition was 2.6008. A three-cell loop with product 309.494 set the bound. One kernel had a weighted signed integral only 0.34% of its absolute integral. An added cancellation filter removed four pairs and lowered the factor to 1.3881. A stronger filter left 66 pairs and a factor of 1.3088. These filters were added after the main result. Four passive controls retained all 302 neuron slots and every electrical pair in four separate anatomical exports. They passed with unequal leaks, capacitances and cell filters. We preserve transforms for all 9,540 nonzero WT kernel entries, 10,231 neuronal graph facts, code, certificates, figures and repeat steps. Exact rational arithmetic confirmed all 424 primary transform values. Fresh source downloads reproduced eight numerical input files; an offline repeat matched 57 result files. The contribution is a dynamic reciprocity audit and a measured response-cancellation effect. The bounds describe fixed fitted values. They are not confidence limits or proof of nonreciprocal physiology. The export lacks event-level kernel fits and animal IDs, and conditional averages from changing states need not follow one reciprocal model.
C. elegans · connectome · signal propagation · reciprocity · calcium kernel · Laplace transform · cell filters · response cancellation · gap junction · reproducibility
References
10- usesdoi:10.1038/s41586-023-06683-4
Original response measurements and fitted kernels.
- usesdoi:10.17605/OSF.IO/E2SYT
Original functional data record. The export is pinned separately by commit and hash.
- usesdoi:10.1098/rstb.1986.0056
White whole and L4 neuronal graph exports.
- usesdoi:10.1038/s41586-021-03778-8
Two adult neuronal graph exports, with all neuronal relays retained.
- usesdoi:10.1007/s10107-010-0419-x
Established graph cycle and cell-potential decomposition.
- supportsdoi:10.1007/s00422-016-0681-y
Electrical network response functions and their relation to reciprocal transfer.
- usesWith free cell gains, changes within 1.25 sample standard errors allowed a reciprocal fit to 109 selected C. elegans response pairs sci:sha256:cd8e2262408cdd13ad3f6faa138f2b822a795ac9f993a1571ead9954cd16076a
Prior scalar-gain reciprocity audit of worm mean responses. We use its audited source scope and independently reproduce all native neuronal graph facts. We add a dynamic filter and response-cancellation audit.
- supportsCycle components of SEEG pulse size ratios were invariant to fixed scalar site gains in six sessions sci:sha256:730f1c79a2cacd4597fa7f3961c4db1c917d7c75ca8bf8bd07342218f7993c63
Earlier scalar site-gain cycle audit of human stimulation responses.
- supportsAnatomical features gave no clear added shape prediction gain in a fixed ridge test of 686 C. elegans transfer kernels sci:sha256:cbf5601ea590f4aa83137cc12fde16e083e067cac724ca2680211ee6cfa8ad4e
Prior common-input kernel-shape benchmark on the same fixed worm export.
- supportsIn a 30-second window, 315 of 668 selected C. elegans kernels with peaks before one second had most absolute area after one second sci:sha256:1e50fd4224cd00f6660aaf9f078f8be64a368a1b56525e5d355277ba0c49c955
Prior kernel peak and tail audit. It establishes related timing context, not the new dynamic cycle result.
Cited by
6 publicationsPublic catalog entries on Sciscape.
- A one-second source pulse shift gave opposite signs in 202 of 739 fitted C. elegans transfer models under a 10% kernel change bound
This earlier study measured cancellation in signed kernel integrals. We independently recover its 739-direction kernel scope before reciprocal pairing.
- supportsTwo C. elegans pharynx response matrices had minimum completion rank two at a Laplace rate of 1 per second
Shows why cancellation and pooled fitted functions need separate checks. We add a time-domain cancellation filter and full-graph state-pooling rank controls.
- usesFor one exact C. elegans mean kernel pair at 0.2 per second, a passive bound required 2.24 fold readout variation with free pools and 4.24 fold with equal pools
It supplied the pinned dynamic kernel transforms and cancellation audit used by the later passive bound study. We verify the source coefficients again.
- usesAt 0.2 per second, six of 66 fitted C. elegans kernel pairs exceeded a passive return bound that holds after averaging with fixed cell filters
We reuse verified numerical transform, absolute-integral and native-graph facts. We add a pooling-preserved source-normalized cycle upper bound and independent source checks.
- With unrestricted positive cell gains at DC, mean responses from one fixed connected passive gap model can match any positive matrix
This study tests dynamic reciprocity and notes that means of different states need not follow one reciprocal model. We give exact averaging results with fixed and changing gains.
- usesAt five Laplace rates, a stricter cancellation filter reduced log convexity failures from 34 of 60 to 4 of 28 fitted C. elegans kernel pairs with products below one
This study supplied the original derived kernel transforms and cancellation facts through the later passive cycle package.