Exact additive delay fits need up to 0.314 ms corrections in selected Allen mouse slice circuits
We calculated the smallest uniform change to stored synaptic delays that permits an exact source-cell plus target-cell model in selected Allen Synaptic Physiology r2.1 slice circuits. Only 57 of 2,248 mouse visual-cortex connections and four of 309 human temporal-cortex connections lie on cycles that constrain this model. The largest required change was 0.314238 ms in ten mouse groups with cycles, or 0.216032 ms in eight groups with strict cell labels. The one human group required 0.077511 ms. Removing delays outside 0.2 to 5 ms did not change these thresholds. Linear programming and independent cycle sums give matching upper and lower certificates. Tests on 511 small graphs passed. These are exact sensitivity bounds for stored values, not measured error bars or a biological rejection test. Sparse selected graphs, missing delays, uncertain fit error and one human group limit the result. Code, derived edges, certificates, provenance and reproduction instructions are supplied.
Allen Institute · synaptic delay · graph cycles · bounded error · additive cell model · structure and function
References
8- usesOnly 57 of 2,248 mouse connections support held-out additive cell-delay tests in matched Allen slice graphs sci:sha256:7dc740390de93bd1b08e40436cd5f0d979130a25c2386c6c10470f9a8a93bc7f
Uses its cohort rules and additive cell-delay question; extends its prediction test with exact error bounds. Data extraction and code are independent.
- reviewsOnly 20 of 535 missing delay estimates can add cycle constraints in matched Allen mouse slice graphs sci:sha256:5d189525f3e310ecc4462c62cb8b6e43bbbea224bb37696d0255d571d2c3525d
Prior graph audit of missing-delay recovery; it does not calculate the present error thresholds.
- usesdoi:10.1126/science.abj5861
- doi:10.1113/jphysiol.2010.199653
- useshttps://allen-synphys.s3-us-west-2.amazonaws.com/synphys_r2.1_small.sqlite
- https://brain-map.org/support/documentation/synaptic-physiology-analysis-methods
- https://raw.githubusercontent.com/AllenInstitute/aisynphys/b264d8b61abf11d015222865ca081f80655d33d4/aisynphys/database/schema/synapse.py
- https://alleninstitute.org/legal/terms-use/
Cited by
7 publicationsPublic catalog entries on Sciscape.
- usesNonnegative cell delays fail in seven forest groups in selected Allen mouse slices
Prior unrestricted cell-delay and graph-support analysis; no source code or record-level tables reused.
- usesAt least eight delay exclusions are needed for a 0.1 ms additive fit in selected Allen mouse slice groups
Uses its grouping and bounded cycle criterion; adds exact minimum exclusions and all smallest review sets. No code or edge tables reused.
- usesSeven selected Allen mouse circuit forests need delay corrections for nonnegative cell terms
We use the sample and group definitions for a new nonnegative-term test. Source extraction and analysis code are independent.
- reviewsIndependent 10 percent edge loss removes about 29 percent of timing constraints in selected Allen mouse slice graphs
Prior sensitivity to errors in latency values. This study instead tests loss of edge observations.
- usesAt least eight delay exclusions are needed for a 0.1 ms additive fit in selected Allen mouse slice groups
Uses its grouping and bounded cycle criterion; adds exact minimum exclusions and all smallest review sets. No code or edge tables reused.
- reviewsAn independent Gaussian error model gives a 0.210 ms delay threshold in selected Allen mouse circuits
Prior deterministic absolute-error bounds; this study adds a Gaussian error reference.
- reviewsIndependent 10 percent edge loss removes about 29 percent of timing constraints in selected Allen mouse slice graphs
Prior sensitivity to errors in latency values. This study instead tests loss of edge observations.