ISN Detection: Model Predictions vs Experimental Evidence Across Cortical Areas

The debate over ISN detection reveals that two-population models vs four-population models make fundamentally different predictions about what paradoxical effects mean for cortical architecture. Layer-specific analysis further complicates the picture, suggesting ISN may not be a uniform property of cortex.

{
  "papers": [
    {
      "n": 0,
      "doi": "10.7554/elife.54875",
      "value": "3",
      "method": "two-population ISN model fit to optogenetic data",
      "metric": "Cortical areas showing ISN operation",
      "n_analyzed": null,
      "ci_or_error": null,
      "text_access": "fulltext",
      "n_definition": "cortical areas tested",
      "scope_region": "V1, S1, motor cortex",
      "study_system": "mouse cortex in vivo",
      "taxonomic_level": "PV interneurons",
      "scope_population": "PV+ interneurons and pyramidal cells",
      "value_source_sentence": "We find clear evidence for ISN operation in mouse visual, somatosensory, and motor cortex.",
      "experimental_conditions": "optogenetic perturbation of PV+ neurons, transgenic expression"
    },
    {
      "n": 0,
      "doi": "10.1523/JNEUROSCI.0963-17.2017",
      "value": "large proportion",
      "method": "multi-neuron network simulation",
      "metric": "Fraction of inhibitory population needed for ISN detection",
      "n_analyzed": null,
      "ci_or_error": null,
      "text_access": "fulltext",
      "n_definition": "model neurons perturbed",
      "scope_region": "neocortex (generic)",
      "study_system": "computational cortical networks",
      "taxonomic_level": "all inhibitory neurons",
      "scope_population": "inhibitory neurons",
      "value_source_sentence": "Our models predict that a large proportion of the inhibitory network must be perturbed to reliably detect an ISN regime robustly in cortex.",
      "experimental_conditions": "simulated optogenetic perturbations"
    },
    {
      "n": 0,
      "doi": "10.1016/j.neunet.2023.07.020",
      "value": "2 of 3 layers (L4, L5 but not L2/3)",
      "method": "neural mass model with layer-specific connectivity",
      "metric": "Cortical layers in ISN regime",
      "n_analyzed": null,
      "ci_or_error": null,
      "text_access": "fulltext",
      "n_definition": "cortical layers modeled",
      "scope_region": "cortical column (all layers)",
      "study_system": "multi-layer cortical column model",
      "taxonomic_level": "inhibitory populations (not PV-specific)",
      "scope_population": "excitatory and inhibitory populations per layer",
      "value_source_sentence": "The results show that layer 2/3 in the model does not operate in the ISN regime but layers 4 and 5 do operate in the ISN regime.",
      "experimental_conditions": "Allen Institute-derived connectivity"
    },
    {
      "n": 0,
      "doi": "10.7554/elife.49967",
      "value": "not evidence for ISN",
      "method": "four-population analytical model and simulations",
      "metric": "Interpretation of paradoxical response",
      "n_analyzed": null,
      "ci_or_error": null,
      "text_access": "fulltext",
      "n_definition": "network model populations",
      "scope_region": "ALM L2/3 and L5, S1",
      "study_system": "mouse ALM and S1 cortex",
      "taxonomic_level": "PV, SOM, VIP subtypes",
      "scope_population": "PV, SOM, VIP interneurons and pyramidal cells",
      "value_source_sentence": "In both four-population models, the paradoxical effect implies not too strong recurrent excitation. It is not evidence for stabilization by inhibition.",
      "experimental_conditions": "PV photostimulation in vivo"
    }
  ],
  "comparison_id": "isn-detection-methods-predictions",
  "comparison_name": "ISN Detection: Model Predictions vs Experimental Evidence Across Cortical Areas",
  "comparison_type": "cross-study conflict",
  "what_it_reveals": "The debate over ISN detection reveals that two-population models vs four-population models make fundamentally different predictions about what paradoxical effects mean for cortical architecture. Layer-specific analysis further complicates the picture, suggesting ISN may not be a uniform property of cortex.",
  "homogeneity_check": {
    "caveats": [
      "Mix of experimental and purely computational studies",
      "Different cortical areas and layers compared",
      "Two-population vs four-population models make fundamentally different assumptions",
      "Paper D's 'not evidence for ISN' is a qualitative interpretation, not a quantitative metric"
    ],
    "n_definition_uniform": "false",
    "scope_region_uniform": "false",
    "taxonomic_level_uniform": "false",
    "scope_population_uniform": "false"
  },
  "suggested_plot_type": "grouped bar"
}

[
  "paper:paper-0f05fb84ee59",
  "paper:paper-38fe71bdf9b6",
  "paper:paper-5135d667ceac",
  "paper:paper-d6edd2b57842"
]

{
  "mode": "public_source_pointer_with_short_context",
  "notes": [
    "Local review repositories are read-only inputs.",
    "SciDEX stores paper metadata, structured evidence, file pointers, and short citation contexts; it does not copy full review prose."
  ],
  "source_commit_sha": "df9fc7e8d455b084152c9d713558dae0013cef21",
  "source_repository_url": "https://github.com/AllenNeuralDynamics/ComputationalReviewPV"
}