Enregistré dans:
| Auteur principal: | |
|---|---|
| Format: | Recurso digital |
| Langue: | |
| Publié: |
Zenodo
2026
|
| Sujets: | |
| Accès en ligne: | https://doi.org/10.5281/zenodo.18968564 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
Table des matières:
- <h2>Zenodo Description</h2> <p><strong>ENAQT Boundary Conditions on the Allen Mouse Cortex Connectome</strong></p> <p>This dataset accompanies the study testing whether environment-assisted noise-augmented transport (ENAQT) — a hallmark signature of the Connectome Routing Networks (CRN) framework — arises on the directed weighted mouse cortex connectome from the Allen Brain Atlas (34 cortical regions, 17 per hemisphere).</p> <p><strong>Background.</strong> Previous work demonstrated ENAQT (inverted-U selectivity over dephasing rate κ) on human connectome subcortical motifs (basal ganglia → thalamus, 8/8 subjects). The present study asks whether this signature generalizes to dense cortical association networks.</p> <p><strong>Main result.</strong> Zero inverted-U profiles were detected across 37 conditions: 6 fMRI co-activation pattern (CAP) derived target sets × 2 weight normalizations, 1 DMN-vs-LCN biologically motivated task, 8 sparsity thresholds (density 0.30–1.0, diameter 1–4), and 16 single-hemisphere reduced-target configurations. All κ-curves showed monotonic decay — coherent transport was always optimal at minimal dephasing, with no room for dephasing-assisted improvement.</p> <p><strong>Predictor analysis.</strong> Systematic comparison of topological descriptors across 35 graphs (8 HCP positive, 20 Erdős-Rényi, 7 mouse cortex variants) identified an empirical two-dimensional boundary: ENAQT on weighted biological graphs requires algebraic connectivity λ₂ > 1.0 AND weight coefficient of variation CV ∈ (0.2, 0.5). The Allen mouse cortex violates both (λ₂ < 0.7, CV > 0.9). This rule achieves 100% specificity (zero false positives) on all tested weighted graphs.</p> <p><strong>Mechanistic finding.</strong> ENAQT on HCP survives at zero diagonal disorder (ε = 0), confirming it is purely topology-driven. Spectral overlap η₀ and dark-state indicators were evaluated but do not predict ENAQT across graph types.</p> <p><strong>Conclusion.</strong> ENAQT is a property of hierarchical bottleneck circuits (basal ganglia, mushroom body), not of dense cortical sheets. This defines quantitative applicability limits for the CRN framework.</p> <p><strong>Contents.</strong></p> <ul> <li>4 Python scripts (GKSL κ-sweep, sparsity sweep, predictor hunt, dark-state diagnostic)</li> <li>8 output CSVs covering all 37 conditions + 35-graph predictor table</li> <li>7 publication-quality figures (300 dpi)</li> <li>4 execution logs</li> <li>Manuscript draft (Markdown)</li> <li>SHA256 manifest for reproducibility</li> </ul> <p><strong>Source data.</strong> Mouse cortex connectivity and fMRI from Fasoli et al. (2026, PLoS Comput Biol, doi:10.1371/journal.pcbi.1013995), Mendeley doi:10.17632/xscxtshgfx.2. HCP reference data from Dolgikh (2025), Zenodo doi:10.5281/zenodo.18519173.</p> <p><strong>Environment.</strong> Python 3.11, conda environment <code>crn_clean</code> (numpy, scipy, pandas, matplotlib, networkx, openpyxl).</p> <p><strong>Related works.</strong></p> <ul> <li>CRN FCN Paper (rate-matched wave vs classical transport): doi:10.5281/zenodo.18960172</li> <li>CRN HCP ENAQT Core Dataset: doi:10.5281/zenodo.18519173</li> <li>CRN Drosophila DES Dataset: doi:10.5281/zenodo.18697116</li> <li>CRN Energy Analysis (C5c): doi:10.5281/zenodo.18785551</li> </ul>