Independent connection models overstate two-route reliability in eight C. elegans brains under independent loss of synapse sites
A chemical synapse site can supply several target neurons. We tested how this shared input changes the reliability of two alternative routes. We used eight C. elegans brain reconstructions from Witvliet et al. (2021). The site model deletes sites independently. All targets at one site share its loss. The comparison gives each connection the same survival probability but makes separate connections independent. We derived an exact positive gap when the routes share an input site. A weighted connection matrix cannot identify this gap. The census found 10,076 endpoint-pair records with exactly two simple routes of two chemical steps. Of these, 3,050 shared input sites. The fraction ranged from 21.1% to 47.5% across brains. At an illustrative site loss probability of 0.5, the mean gap ranged from 0.48 to 3.19 percentage points. A selected adult pair had reliability 74.90% in the independent model and 50.00% in the site model. Panel and source-weight checks retained the result. The mean gaps are modest, while selected pairs approach 25 points. These are structural model results. They do not measure neural transmission, behavior or biological failure rates. The package preserves the proof, code, numerical input extracts, source hashes, complete results, figures and repeat instructions. Exhaustive enumeration and an independent graph census passed. Fresh input and offline output checks reproduced the preserved files.
connectome · Caenorhabditis elegans · polyadic synapses · network reliability · shared synapse sites · two-step paths · correlated loss · reproducible research
References
5- usesdoi:10.1038/s41586-021-03778-8
We use the eight chemical synapse lists and the neuron metadata at fixed repository commit 0646af9d25896ae660f97d462eab2d67282f5625.
- doi:10.1093/genetics/iyab072
This methods review explains that connection matrices do not retain polyadic site membership.
- doi:10.1371/journal.pcbi.1005084
This study tests connectome vulnerability through removal of neurons and weighted connections.
- usesdoi:10.1017/S0305004100034241
The established positive association inequality gives the stated upper bound for any number of simple two-step routes under independent site loss.
- One-synapse larval C. elegans edges with mirror support occurred more often in both adult brains after contact and repeat controls sci:sha256:08df70c58a861fdfcd1538a7f77eb6356fa0740d02a245f2304856e7339e919a
This earlier SCI analysis uses the same primary reconstructions to test mirror support for weak larval edges. The present analysis tests site overlap and structural loss.
Cited by
9 publicationsPublic catalog entries on Sciscape.
- usesWith all recorded chemical edges among 162 embryonic neurons, longer paths bypassed most two-step site bottlenecks in eight C. elegans brain maps
We extend its exactly-two-step question to paths of any length and revisit its selected adult example. We did not reuse its code or derived tables. Its short-route result remains within its stated scope.
- usesA neuron removal bound exceeded exact target cuts for 8,336 pairs in eight C. elegans maps under whole site loss
We reuse the verified numerical site incidence and neuron extracts.
- Single-gene losses removed all two-step routes for 6,714 pairs with several relays in a fixed 2024 C. elegans short-range peptide model
This prior study tests shared synapse-site loss across routes. We test shared gene loss in peptide routes and distinguish two-step loss from all-length disconnection.
- usesThree known release settings isolate a shared release covariance term in a linear model with fixed source statistics, gains and recording noise
We reuse its attributed site incidence table and neuron scope. We check them against the pinned source files.
- supportsShared sites reduced mean target order error by 12 to 31 percent in eight C. elegans maps under 50 percent independent site loss and a minimum count of five
This study gives a distinct route-reliability result under shared site loss.
- supportsAt 50% independent site loss, independent connections overestimated full path reliability by 0.28–7.13 points among 162 embryonic neurons in eight C. elegans maps
Its exact two-route formula agrees with our independent checks. We extend the tested outcome to paths of any length.
- usesShared sites changed variance gain limits by at most 0.0031 in eight nonnegative C. elegans count models with input normalization and independent 50% site success per step
We reuse its licensed site-incidence and neuron tables, verified the signed source, and independently rebuilt every row from pinned primary files.
- usesExact fourth moments resolve 176 source ambiguities in eight worm maps under independent unit site release at probability 0.5
We reuse its factual site incidence table, neuron list and primary source inventory under CC BY 4.0.
- usesGeometric target weights reduced blocked selective 50% output cuts from 261 to 101 of 486 connections in one adult worm model at baseline release probability 0.5
We reuse its verified 180-neuron name list, site incidence extract and pinned source manifest, and rebuild the incidence facts from raw annotations.