Recorded movie similarity gave no clear loss decrease after model similarity in a fixed test of 234,393 MICrONS pairs from one mouse
We tested if recorded movie response similarity adds connection prediction information after visual-model similarity and anatomy are known. We used the MICrONS functional connectomics v1 release from one mouse. The test retained 234,393 ordered pairs from 144 reconstructed source neurons to 8,905 targets. There were 6,608 connections. Four fixed logistic models used common anatomy and response quality controls. Five folds held out complete source groups. Adding recorded similarity after model similarity reduced log loss from 0.115653412 to 0.115610514 nats per pair. The decrease was 0.037%. The central 95% source and target resample range was -0.0000552 to +0.0001443 nats per pair. Positive values mean lower loss. The range included zero in target spatial holdouts, linear function terms, and two penalty checks. A clean repeat gave identical test probabilities. This result concerns conditional prediction in one selected anatomical panel. It does not establish equal biological information, absent functional wiring, or causal signal flow. The package contains the complete scalar opportunity table, source hashes, fixed plan, code, predictions, scores, figures and repeat steps.
connectome · MICrONS · structure and function · recorded responses · digital twin · connection prediction · group holdout · data provenance
References
6- usesdoi:10.1038/s41586-025-08840-3
We use the released node and edge tables. They give anatomical opportunities and both recorded and model signal correlations.
- usesdoi:10.1038/s41586-025-08790-w
This work supplies the MICrONS functional and anatomical resource from one mouse.
- doi:10.1038/s41586-025-08829-y
This work describes the visual response model used to estimate functional properties.
- usesIn the V1 Deep Dive mouse, connected excitatory neurons had only slightly more similar movie responses than nearby unconnected neurons; the same test gave a 2 to 7 times larger effect in MICrONS sci:sha256:dd15d6d8156d77ee6c2bf3f36180346ce5d8f7d0ceecd54d2e329703851a4469
We reuse its cached publisher files and build on its comparison of empirical function and anatomy. We independently check the files against the publisher. We do not reuse its analysis code.
- Excitatory axons, but not basket cell axons, selected excitatory targets with similar orientation beyond axon–dendrite overlap and 10 µm soma distance in MICrONS mouse visual cortex sci:sha256:3475d3289a96669e6e83d55e0fc9ec509894213422121532387c543739c8dca2
This earlier SCI study tested model orientation specificity after anatomical overlap controls. Our test uses recorded and model movie similarities together.
- In MICrONS mouse visual cortex, connected excitatory neurons had more similar preferred orientations mainly when their synapses were large (14,412 neurons, digital-twin tuning) sci:sha256:07b8a9b83eea2b7443dc99d29805440aefc0d23414bb9c8959decc534824bd92
This earlier SCI study tested model orientation matching and synapse size. It helps define the different scope of the present conditional prediction test.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.