Skip to content

DD027 Draft GitHub Issues

Epic: DD027 — Multicompartmental Neuron Models

Generated from: DD027: Multicompartmental Neuron Models

Methodology: DD015 §2.2 — DD Issue Generator

Totals: 3 issues (ai-workable: 1 / human-expert: 2 | L1: 1, L2: 0, L3: 2)

Roadmap Context: DD027 is a Phase 2 DD (proposed). These issues were originally part of DD001 Draft Issues (Groups 5 and Infrastructure) and have been relocated here because multicompartmental modeling is now specified by DD027.

Group Phase Rationale
1. Level D Multicompartmental (Issues 1-2) Phase 2 Proof-of-concept requires EM morphologies + Nicoletti channels
2. Infrastructure (Issue 3) Phase A Config toggle depends on DD013 openworm.yml

Group 1: Level D Multicompartmental Development (Phase 2)

Target: Multicompartmental neuron models for neurons where single-compartment approximation is insufficient.

Issue 1: Evaluate and refine existing NeuroML2 morphologies for Level D neurons

  • Title: [DD027] Evaluate existing CElegansNeuroML morphologies for AWC, AIY, AVA, RIM, VD5 and refine for Level D
  • Labels: DD027, human-expert, L3
  • Roadmap Phase: Phase 2
  • Target Repo: openworm/c302
  • Required Capabilities: python, neuroml, neuroanatomy
  • DD Section to Read: DD027 — Implementation Pathway (Stage 1, step 2)
  • Depends On: DD005 Issue 5 (channel survey, which also catalogs morphology sources)
  • Existing Code to Reuse:
    • CElegansNeuroML/CElegans/generatedNeuroML2/ — All 302 neurons already exist as multicompartmental NeuroML2 cells with realistic 3D morphology (derived from VirtualWorm Blender files). Includes AWCL.cell.nml, AWCR.cell.nml, AIYL.cell.nml, AIYR.cell.nml, AVAL.cell.nml, AVAR.cell.nml, RIML.cell.nml, RIMR.cell.nml, VD5.cell.nml — all 5 target neurons.
    • BAAIWorm/eworm/components/model/ — HOC files for every neuron with multi-compartment morphology and biophysical parameters. Includes per-neuron conductance JSONs specifying which channels are expressed where.
    • c302/NeuroML2/ — Another copy of all 302 neuron morphologies already in the c302 repo
  • Approach: Evaluate existing morphologies, do not recreate. The NeuroML2 morphologies for all 5 target neurons already exist in two repos. Evaluate whether they meet Level D requirements (segment length < 2µm, segment group annotations). Compare against BAAIWorm HOC files for discrepancies. Refine as needed.
  • Files to Modify:
    • morphologies/AWC.cell.nml (refined from existing CElegansNeuroML, if needed)
    • morphologies/AIY.cell.nml (refined)
    • morphologies/AVA.cell.nml (refined)
    • morphologies/RIM.cell.nml (refined)
    • morphologies/VD5.cell.nml (refined)
  • Test Commands:
    • for f in morphologies/*.nml; do jnml -validate $f; done
  • Acceptance Criteria:
    • [ ] Start from existing CElegansNeuroML morphologies — do NOT recreate from SWC
    • [ ] Compare existing NeuroML2 morphologies against BAAIWorm HOC files; document discrepancies
    • [ ] Segments < 2 µm length (adequate spatial resolution for cable equation); refine if existing segmentation is too coarse
    • [ ] Segment groups defined: soma, axon, dendrite (where applicable)
    • [ ] All pass jnml -validate
    • [ ] Can be loaded by NEURON simulator (via pyNeuroML export)
    • [ ] Soma diameter matches known values (WormAtlas)
  • Sponsor Summary Hint: Here's the surprise: multicompartmental NeuroML2 morphologies for all 302 neurons already exist in CElegansNeuroML and are already copied into c302. This issue evaluates whether those existing morphologies are detailed enough for Level D cable-equation simulation, compares them against BAAIWorm's independently derived morphologies, and refines where needed. Evaluation, not creation.

Issue 2: Integrate existing components into AWC Level D proof-of-concept

  • Title: [DD027] Integrate existing Nicoletti AWCon channels + CElegansNeuroML morphology into c302 Level D AWC proof-of-concept
  • Labels: DD027, human-expert, L3
  • Roadmap Phase: Phase 2
  • Target Repo: openworm/c302
  • Required Capabilities: python, neuroml, electrophysiology
  • DD Section to Read: DD027 — Implementation Pathway (Stage 1, steps 3-4) and Reference 3 (Nicoletti et al. 2019 AWCon model)
  • Depends On: Issue 1 (AWC morphology evaluation), DD005 Issues 6-9 (adopted channel library)
  • Existing Code to Reuse:
    • NicolettiEtAl2019_NeuronModels/NeuroML2/AWCon.cell.nml — Complete AWCon single-compartment model with ALL 16 channels in NeuroML2, validated against XPP original. Includes CaDynamics.nml for calcium concentration dynamics and GenerateNeuroML.py for programmatic cell generation.
    • CElegansNeuroML/CElegans/generatedNeuroML2/AWCL.cell.nml — AWC multicompartmental morphology
    • c302/parameters_D.py — Level D framework already exists with multicompartmental infrastructure, ChannelDensity support, Species (Ca), and FixedFactorConcentrationModel. Currently uses "BlindGuess" placeholder channels (generic leak + k_slow). The framework is ready — just needs real channel models plugged in.
    • BAAIWorm/eworm/components/param/cell/ — Per-neuron JSON with max conductance for each channel — use AWC entry for channel density distribution guidance
  • Approach: Integration, not creation. The channels exist (NicolettiEtAl2019). The morphology exists (CElegansNeuroML). The Level D framework exists (c302 parameters_D.py). This issue combines them: distribute NicolettiEtAl2019's 16 AWCon channel models across the existing AWC morphology, using BAAIWorm's conductance JSONs for distribution guidance, within c302's existing Level D framework.
  • Files to Modify:
    • c302/c302_MultiComp_AWC.py (new — or extend existing Level D template)
    • tests/test_awc_multicomp.py (new)
  • Test Commands:
    • jnml -validate morphologies/AWC.cell.nml
    • jnml LEMS_AWC_test.xml -nogui
    • pytest tests/test_awc_multicomp.py
  • Acceptance Criteria:
    • [ ] AWC multicompartmental model built by combining existing NicolettiEtAl2019 channels with existing CElegansNeuroML morphology — not reimplemented
    • [ ] Per-segment channel densities assigned from adopted channel library (DD005 Issues 6-9), guided by BAAIWorm conductance data
    • [ ] Passive parameters (axial resistance, membrane capacitance) fitted to match AWC electrophysiology
    • [ ] Reproduces published AWC responses (Nicoletti et al. 2019) within ±15%
    • [ ] Simulates in NEURON via pyNeuroML export
    • [ ] Can coexist with Level C1 single-compartment neurons in the same network
    • [ ] Documents which existing components were combined and any modifications made
  • Sponsor Summary Hint: The channels exist (16 in NeuroML2 from Nicoletti). The morphology exists (CElegansNeuroML). The Level D framework exists (c302 parameters_D.py with placeholder channels). This issue plugs real channels into a real morphology in an existing framework — proving that Level D multicompartmental neurons work. Three existing codebases, one integration task.

Group 2: Infrastructure (Phase A)

Issue 3: Add neural.spatial_synapses config toggle for Level D

  • Title: [DD027] Add neural.spatial_synapses config toggle for spatially resolved synapse placement
  • Labels: DD027, ai-workable, L1
  • Roadmap Phase: Phase A
  • Target Repo: openworm/c302 + openworm/OpenWorm
  • Required Capabilities: python, yaml
  • DD Section to Read: DD027 — Spatially Resolved Synapse Placement
  • Depends On: DD013 Issue 1 (openworm.yml schema)
  • Files to Modify:
    • openworm.yml (add neural.spatial_synapses: false)
  • Test Commands:
    • python3 -c "import yaml; c = yaml.safe_load(open('openworm.yml')); print(c['neural']['spatial_synapses'])"
  • Acceptance Criteria:
    • [ ] neural.spatial_synapses: false → synapses are abstract neuron-to-neuron (current behavior)
    • [ ] neural.spatial_synapses: true → placeholder for spatially placed synapses (requires Level D)
    • [ ] Config validation: spatial_synapses: true requires level: D (error otherwise)
    • [ ] Documented with DD cross-reference
  • Sponsor Summary Hint: In the simple model, a synapse is just "neuron A connects to neuron B." In the detailed Level D model, synapses have specific locations along the neurite — and location matters because it determines how signals combine. This config toggle enables spatial synapse placement when Level D is active. The actual placement algorithm is a separate issue.

Summary Statistics

Category Count
Total Issues 3
ai-workable 1
human-expert 2
L1 1
L2 0
L3 2
Group Issues Target
1: Level D Development 1–2 Evaluate existing morphologies, integrate AWC proof-of-concept
2: Infrastructure 3 Config toggle for spatial synapses

Cross-References

Related DD Relationship
DD001 (Neural Circuit) Original source — these issues were extracted from DD001 Draft Issues Groups 5 and Infrastructure Issue 20
DD005 (Cell-Type Specialization) Issues 6-9 (channel library) are prerequisites for Issue 2
DD010 (Validation Framework) Level D validation criteria
DD013 (Simulation Stack) Issue 3 (config toggle depends on openworm.yml schema)
DD017 (Hybrid ML) Parameter fitting backend for Stage 1 Step 4
DD020 (Connectome Data Access) Morphology data access
DD024 (Validation Data Acquisition) Synapse centroid distance data

Dependency Graph

DD005 Issue 5 (channel survey)
  └→ Issue 1 (evaluate existing morphologies)
       └→ Issue 2 (AWC Level D integration — channels + morphology + framework)

Issue 3 (spatial_synapses config) — depends on DD013 Issue 1