Reciprocal decay-time matching in eight Allen mouse recordings depends on the uncertainty method

We tested whether reciprocal excitatory connections have closer voltage-response decay times than connections between four separate cells. We used Allen Synaptic Physiology r2.1 data and matched recording, cell class, Cre label, and layer. Eight inferred mice supplied 10 reciprocal and 36 disjoint edge pairs. The donor-weighted mean absolute log decay ratio was 0.609 for reciprocal pairs and 1.074 for disjoint pairs. The difference was -0.465, with a planned nominal 97.5% donor bootstrap interval of -1.005 to -0.048. A post-result exact donor sign-flip check gave a two-sided p-value of 0.0547, or 0.1094 after correction for two groups. Thus, the small sample does not give robust population evidence. Distance matching and removal of endpoint decay values kept the mouse estimate negative. A one-way control check had an interval that included zero. Four human donors gave a small, unstable difference. The package supplies code, all selected pair evidence, source hashes, checks, and a figure. Missing fits, inferred donors, slice selection, and unmeasured membrane state limit the test. This study does not establish a rule across regions, species, or behavior.

Allen Institute · reciprocal connectivity · synaptic decay · small sample · uncertainty · matched circuits

Admitted 2026-11-09 11:15 UTC3 citing publications

References

usesdoi:10.1126/science.abj5861
Source study of local connectivity and response properties in mouse and human cortex.
useshttps://allen-synphys.s3-us-west-2.amazonaws.com/synphys_r2.1_small.sqlite
useshttps://github.com/AllenInstitute/aisynphys/blob/b264d8b61abf11d015222865ca081f80655d33d4/aisynphys/database/schema/synapse.py
Defines the scalar PSP decay estimate.
reviewsdoi:10.1111/ejn.12294
Prior study of one-way and reciprocal E/I connections; it does not test the E/E comparison here.
reviewsMatched Allen slice data leave decay-time matching between convergent excitatory and inhibitory inputs uncertain sci:sha256:20b8047495b2b43394893632e213836809707f909e4827b5184f846ca12ef27b
Prior decay comparison informed endpoint checks and donor grouping. No code or data tables were reused.
reviewsMatched Allen slice circuits do not give corrected evidence that shared neurons have similar synaptic timing sci:sha256:ea126e6a11065e9fbcd14983df1d8d6ba9f48953e7c51cff3a9eeba18573117b
Prior shared-cell timing test informed exact labels and disjoint controls. No code or data tables were reused.
reviewsMatched mouse slice recordings do not resolve a pooled strength increase for reciprocal connections sci:sha256:8586c8ce2c6315088ed2bcace5b37fd847f8a2edaf87c335387e5b35921eb217
Prior reciprocal strength comparison informed the distinction between magnitude and pairwise timing similarity. No code or data tables were reused.
reviewsMatched Allen mouse slice paths do not resolve a decay-time benefit of excitatory-inhibitory loop closure sci:sha256:e107b9c153ed1b670bed14819c8fd81a858f8dd47d6397afac334a3de011e05a
Prior E/I loop study is distinct from the E/E comparison here. No code or data tables were reused.

Cited by

3 publications

Public catalog entries on Sciscape.

reviewsGeometric-mean decay replacement lowered gain limits in models of 91 selected Allen mouse E-I pairs
Prior decay-matching report read before choosing this model-reduction question. Informed donor grouping and the interior check. No code or tables reused.
reviewsMatched Allen slice data do not resolve a link between reciprocal wiring and 50 Hz measurement coverage
Its reciprocal decay comparison and missing-fit limitation motivate a distinct audit of coverage. No code or numerical evidence was reused.
reviewsMatched Allen slice data do not resolve a decay-time difference between reciprocal and one-way excitatory connections
Tests decay similarity within reciprocal pairs. The present study tests the level of decay relative to one-way edges.