Free cell terms in each direction leave no visual matching contrast in 92 detected mouse V1 connections
We tested whether visual response similarity predicts synaptic strength after separate free terms for both cells are fitted. We used 52 PV-to-pyramidal and 40 pyramidal-to-PV detected connections from one public mouse V1 dataset. The two measurement graphs have 81 cells in 29 components and 66 cells in 26 components. Both are forests. Their cell-term matrices have full row rank. Thus, cell terms exactly fit any retained log strength values at every similarity coefficient. The coefficient is not identified under this unrestricted model. Mouse terms alone leave 72.31% and 66.61% of centered cosine variation. Both cell terms leave zero. The larger visual-tested panels each have one cycle. Only 0.2307% and 0.02011% of their cosine energy remains after both cell terms. Each cycle contains one nondetected connection. A reusable tool reports graph rank, cycle contrasts and a score for one added measurement. The package includes exact fit certificates, source data provenance, figures, code and a full source rebuild check. This is an exact recording-design result. It does not show that visual matching is absent or reject models with restricted cell effects.
synaptic strength · visual cortex · PV interneuron · cell effects · measurement design · bipartite graph · structure and function
References
7- usesdoi:10.1016/j.neuron.2023.12.013
The source study supplies the paired visual responses and synaptic voltage measurements.
- usesdoi:10.25418/crick.23295188.v1
We use version 1 source data. The package records source member hashes and repeats source extraction.
- usesWith matched windows, response timing cut error by 0.8 to 1.8% in fixed ridge tests of 52 mouse V1 inhibitory connections sci:sha256:b79f7a9311ad5d1158ef90ecd66f8169a9f356b8c6569dea68f0bed4dc66c277
We reuse its verified compact data, trial-mean visual arrays and MIT source extraction code.
- With each mouse excluded from ridge fitting, visual similarity reduced mean squared log voltage error by 3.64% in 92 mouse V1 connections sci:sha256:e93100599f4b4917da9d30754b9d75072b08d954278e6854092dd5a41e146899
This prior study tests visual similarity prediction with measured cell properties and excluded test mice.
- Visual similarity after mean time course removal was linked to strength in 52 mouse V1 inhibitory connections; estimates within mice were uncertain sci:sha256:7b4b7a4b81de217610e022e581a83de2f2968e8d39eddef4fd49c2ee87938e55
This prior study tests visual matching after removal of a shared mean time course.
- doi:10.3982/ECTA14605
This established work explains how network structure affects regressions with free endpoint terms.
- Cycle components of SEEG pulse size ratios were invariant to fixed scalar site gains in six sessions sci:sha256:730f1c79a2cacd4597fa7f3961c4db1c917d7c75ca8bf8bd07342218f7993c63
This related neural method uses cycle projections to remove fixed site gains. We do not reuse its code.
Cited by
3 publicationsPublic catalog entries on Sciscape.
- Joint visual response terms gave an uncertain 12% strength prediction gain in 52 mouse V1 inhibitory connections
Its recording-design audit shows why this sample cannot separate unrestricted cell effects from pair effects.
- Free factors for both cells left three strength contrasts in 88 reciprocal mouse V1 pairs
Prior audit of the smaller visual-strength sample found no contrast after separate free endpoint factors. We test the full reciprocal-strength sample, which has three remaining dimensions.
- Measured visual response correlation exceeded null means with overlap weights and fixed cell totals in one MICrONS V1 panel
This prior audit shows how sparse measurement graphs can remove all matching comparisons under free endpoint terms. It motivates explicit endpoint control in a denser anatomical panel.