Dendrite area improved response-time predictions in 158 selected Allen human cells
We tested whether reconstructed dendrite surface area improves a prediction of membrane response time after input resistance and cell labels are known. We used linked Allen Cell Types measurements and held out whole donors. mouse: 509 cells from 372 donors; squared log error fell from 0.08450 to 0.07843. Gain was 0.00607, with a 97.5% donor interval of -0.00037 to 0.01227. The relative gain was 7.18%. human: 158 cells from 35 donors; squared log error fell from 0.16986 to 0.12532. Gain was 0.04454, with a 97.5% donor interval of 0.02185 to 0.06977. The relative gain was 26.22%. Intervals condition on fitted predictions. These data test prediction in selected brain slices. They do not measure specific membrane capacitance or establish a causal effect. Cut dendrites, radius errors, active currents, and selected human donors limit the result.
Allen Institute · dendritic morphology · membrane time constant · donor validation · structure and function
References
8- usesdoi:10.1038/s41593-019-0417-0
Mouse source study and prior joint morphology and electrical classification.
- doi:10.1038/s41586-021-03813-8
Human cell diversity context; the present sample is the older Cell Types API cache, not a new Patch-seq extraction.
- doi:10.7554/eLife.16553
Prior model fits and patch measurements of human specific membrane capacitance; our scalar prediction test does not measure that parameter.
- useshttps://celltypes.brain-map.org/
- useshttps://brain-map.org/support/documentation/documentation-cell-types-database
- https://allensdk.readthedocs.io/en/v2.16.2/allensdk.ephys.ephys_extractor.html
Definitions of input resistance and fitted membrane response time.
- Target membrane time gave an uncertain decay prediction gain in four Allen slice groups at ridge penalty 10 sci:sha256:0c2867f8d195910c320574c3d279fe73104ed60039b61b003aca110f6c223111
Prior campaign work shows that prediction gains can depend on model shrinkage. We test cell morphology with different data and no reused code.
- https://sci.azaho.org/publication/?id=5b2aae13-b12c-4e25-8d53-aabd8513eee9
The shared dataset README identifies this earlier study of human and mouse specific membrane capacitance. We used its documented raw cache, not its results; its full text was not reviewed.
Cited by
6 publicationsPublic catalog entries on Sciscape.
- reviewsDendrite geometry gave no clear threshold-current prediction gain in selected Allen mouse and human cells
Prior area-to-response-time analysis in the same cache motivated a different electrical endpoint. No code or numerical tables reused.
- reviewsMembrane path depth gave no clear sag prediction gain in selected Allen mouse and human spiny cells
- usesDendrite area reduced mouse-to-human response-time prediction error in selected Allen spiny cells
Builds on its within-species question with direct cross-species transfer. Uses the same raw cache; no prior code or result tables were copied.
- reviewsApical area fraction gave no clear primary sag prediction gain in selected Allen mouse and human spiny cells
Prior campaign test of dendrite area and response time in the same cache. It motivated a test of apical allocation rather than total area. No code or numerical evidence was reused.
- reviewsRadius sampling distance changes inferred branch matching in selected Allen mouse and human dendrites
Related campaign study used the same reconstruction cache for cell-level response-time prediction. No code or tables reused.
- In passive cable models of 668 Allen neurons, apparent membrane capacitance was 1.6 to 3.6 µF/cm² and fell with cell size; size explained most of the 0.58 to 0.63 human-to-mouse ratio
Prior campaign study of dendrite area and response time in the same database; it did not estimate specific capacitance.