Protean Labs
A provenance-aware scientific operating system for peptide discovery.
Models propose, deterministic validators and review gates decide, and public provenance makes the process inspectable.
The Problem
Scientific AI needs more than generation. It needs evidence, memory, and review. A model can propose candidates, but that is not science.
Science requires that claims trace to evidence, that constraints are enforced, that provenance survives, and that human review gates what gets published.
Protean Labs builds the infrastructure layer that makes this possible.
The System
A layered architecture where each stage has explicit inputs, constraints, and outputs.
Evidence Retrieval
Structured scientific evidence from literature, databases, and prior experiments informs every hypothesis. The system knows what is known.
Hypothesis Formation
Bounded hypothesis generation constrained by evidence. Models propose; constraints shape what gets proposed.
Candidate Generation
Peptide candidate generation within validated constraints. The system produces candidates, not conclusions.
Deterministic Validation
Rule-based validators enforce scientific constraints. Hard rules gate what survives, regardless of model confidence.
Public Provenance
Every claim traces to its computational origin. Provenance trails make the process inspectable and replayable.
Review Gates
Human review before publication or deployment. Automation proposes; review decides what ships.
What This Enables
Infrastructure properties that make scientific AI auditable and trustworthy.
Provenance-Aware
Every output traces to evidence, constraints, and computational steps.
Deterministically Gated
Rule-based validators enforce hard constraints regardless of model outputs.
Review-Required
Human review gates before any output gains authority or publication.
Publicly Inspectable
Public proof surfaces make the scientific process auditable.
Scientific Boundaries
Computational rankings are research-prioritization signals. They do not constitute biological validation, therapeutic claims, or clinical readiness.
Protean Labs produces computationally-ranked peptide candidates for research prioritization. Biological activity, safety, and efficacy require wet-lab validation and are not claimed by computational outputs.
Public proof surfaces expose provenance and computational methodology, not biological proof. The system is research infrastructure, not therapeutic development.
Interested in Protean Labs?
For research collaborations, partnerships, or inquiries about Protean Labs.