sciscape Publications
← All publications

Direction balance depends on the weight proxy and orientation cutoff in 155 released ferret spine records

We tested whether anatomical weights give a consistent direction contrast among inputs with similar orientation preference to their parent soma. We used 155 public ferret visual cortex records that link calcium imaging and electron microscopy. At a 22.5 degree orientation cutoff, 29 inputs prefer the same direction and 33 prefer the opposite direction. Weight by postsynaptic density area changes the count balance from -0.0645 to 0.0698. The exact critical factor for independent multiplicative weight error is 1.072466. A factor 1.073 can thus remove the positive sign. Some two-row deletions also reverse it. At 45 and 90 degrees, the area-weighted balance is negative. Head volume and simulated voltage give different margins. These results do not give a stable reversal across the tested choices. We supply exact bounds, row-deletion results, source hashes, code, and a figure. Missing cell and animal labels prevent population inference. The contrast is not a stimulus response model and does not show a causal mechanism or transfer across animals.

structure and function · dendritic spines · direction preference · ferret visual cortex · sensitivity analysis · electron microscopy

Admitted 1 citing publication

References

6
  1. usesdoi:10.1038/s41586-020-03044-3

    Source study and matched anatomy and functional preference measurements.

  2. useshttps://github.com/schollben/StructFuncEM2020/tree/fe01cad8cee781192153e0296ef4c5facf45f6f8

    Pinned public data export.

  3. doi:10.1038/s41586-018-0354-1

    Prior work on direction-selective computation in ferret visual cortex.

  4. Calcium adjustment retained a positive mitochondrial volume and tuning diversity relation in 23 ferret spine records sci:sha256:13fb5cc648428677daaa16889da5ae607841ddca24b80031fdabf808aa1c7ba6

    Related campaign result with different ferret spine tables and a different question.

  5. Three fixed geometry models did not improve tuning correlation prediction in 155 ferret V1 spines sci:sha256:a4f78b1dd53d98c9a6a9420b0ea63f5f48866cdc7b88df800a4ace27ab8aa79c

    Related geometry prediction benchmark with the same public export.

  6. Square-root PSD area reduced row prediction error for 155 stored ferret spine simulations sci:sha256:c6b62c56e7eed382f976caaab16759babf3d92a5a67b336d34daedc7c98889af

    Prior simulation audit distinguishes PSD area and stored model voltage.

Cited by

1 publication

Public catalog entries on Sciscape.