Skip to content
Protean LabsDocs

Collections

Collections are first-class scientific groupings. They organize candidates, hypotheses, experiments, papers, assay requests, and provenance records into coherent research cohorts.

Uses

  • benchmark sets
  • assay cohorts
  • experiment batches
  • validation groups
  • exploration branches
  • public redacted releases

Lifecycle

private
-> internal review
-> assay pending
-> assay active
-> public redacted
-> public full redacted
-> archived

Most collections remain private or internal review objects. Public collections expose redacted membership, hashes, lifecycle state, and provenance context. Public full redacted means the public-safe metadata for a reviewed collection is complete; it does not mean raw scientific payloads are public.

Collections As Scientific Operating Units

Collections let Protean reason about groups, not just individual candidates. They are where runtime output becomes reviewable research: candidate families, evidence traces, assay readiness, provenance commitments, and commercialization context held together without exposing private payloads.

Collections support wet-lab batch planning, failure-memory studies, exploration branches, public proof surfaces, and future protocol coordination research.

Future Network-Aligned Participation Structures

Collections may also become organizing units for future network-aligned participation research because they bind provenance, scientific object lineage, disclosure state, review milestones, wet-lab progression, and public-safe artifacts.

A collection remains a scientific cohort first. Any future mechanism connected to collection state would be separate from $PRTN itself, separately documented, and subject to legal, regulatory, jurisdictional, compliance, scientific, tax, and operational review.

Runtime Coordination

The verified Hermes workflow definition places collection updates after ranking, claim QA, and provenance preparation. The active loop does not yet schedule collections automatically. The collection builder emits a private registry for local review and a public projection that contains only commitments, redacted object refs, lifecycle state, and integrity hashes.

python3 pipelines/collection_system.py --build
python3 pipelines/collection_system.py --verify

Collection manifests are public-safe by construction: no raw sequences, salts, local paths, embeddings, scoring weights, source traces, or wet-lab automation fields.