External Collaboration Model

The gen5 Pattern

gen5 asks: does someone else’s science come out the other end? The external collaboration model makes this architecturally possible — collaborators use the same infrastructure patterns as internal gates, scoped to their domain.

Sovereignty Enables Collaboration

Sovereignty is not isolation. The ecosystem’s self-hosted infrastructure (Forgejo, WaterFall sync, sovereign DNS) provides the substrate for external collaboration without vendor lock-in:

• Collaborators sync through the same pipelines as internal gates
• Data stays on sovereign infrastructure (no cloud vendor ingestion)
• Provenance is tracked end-to-end (every computational step attributable)
• The collaborator owns their output (pseudoSpore as delivery format)

The Collaborator Gate Model

External collaborators get a gate profile in the ecosystem manifest, scoped to repos relevant to their domain:

[gates.gonzales_nf]
repos = [
    "wateringHole",
    "helixVision", "initioChem", "blueFish",
    "wetSpring", "hotSpring",
    "projectFOUNDATION",
]

They pull only what they need. They never see NUCLEUS internals, unrelated springs, or infrastructure repos. The WaterFall pipeline handles scoping automatically.

What the Collaborator Brings

1. A biological question the ecosystem hasn’t answered
2. Domain data access (NF Data Portal, LTEE datasets, analytical standards)
3. Domain expertise (signaling biology, microbial evolution, analytical chemistry)
4. Institutional authority (PI status for grants, publication lead)

What the Ecosystem Provides

1. Validated computation — 12,510+ checks, 70+ papers reproduced across 8 domains
2. Multi-product composition — orchestrated products for the collaborator’s question
3. GPU compute at zero cost to the institution
4. Self-verifying artifacts — pseudoSpore packaging with full provenance
5. AI-accelerated coordination — metadata extraction, cross-checking, assembly

What Comes Out

1. A pseudoSpore — self-verifying data package the collaborator owns
2. Grant preliminary data — foundation-ready computational evidence
3. New spring validation targets — the ecosystem grows from collaborator science
4. A reproducibility record — every step provenance-tracked

Multi-Product Composition (gen5 novelty)

gen4 products each composed primals independently. gen5 demands products compose with each other — driven by the biological question, not internal design.

Example — neurofibromatosis data mining requires:

• helixVision for gene expression mining from NF Data Portal
• healthSpring for drug repurposing scoring against NF targets
• initioChem for conformational dynamics of inhibitor binding
• coralForge for structural variant impact prediction

No single product answers the question. The answer emerges from their composition — orchestrated by the science itself.

The Spore Cycle

The collaboration completes a biological cycle:

Ecosystem validates published science (springs)
  → Products compose validated computation
  → Collaborator produces new science
  → New science → new validation targets for springs
  → Springs evolve from external demand
  → Ecosystem is stronger than before

Current Collaborators

Scientific Challenges

Beyond individual collaborators, the ecosystem participates in structured benchmarks hosted by scientific foundations (Synapse/DREAM challenges):

• Docker-based submission maps to pseudoSpore pattern
• Foundation-sponsored evaluation builds credibility with funders
• Domain-expert scoring validates products under external criteria
• Challenge results reveal gaps internal testing never surfaces

This is gen5 validation at population scale — not one collaborator’s question, but the entire field’s benchmarks.