CNB-1 is the first experimental model of Project Proteus. The question is whether it can yield concrete protein candidates; so far it has produced two.
Abstract
CNB-1 is Proteus’s first experimental model. This article asks whether it can produce concrete protein candidates and reports the two outputs documented so far.
1. CNB-1 and Proteus
Project Proteus is the broader research program; CNB-1 is the individual model first used to explore AI-assisted protein design.
PROJECT Project Proteus
MODEL Caelis Neural Base-1
ABBREVIATION CNB-1
ORGANIZATION Inserloft Research
ROLE First experimental Proteus model
OUTPUT Computational protein candidates
STATUS Experimental research2. The first documented design
A candidate of about 350 residues, seeded from a PETase-associated motif and aimed at breaking down PET. Its features still await validation.
3. A new CNB-1 output: AGD-F1
AGD-F1 is a 377-residue computational protein candidate generated from a structural seed and modelled with ESMFold.
4. What AGD-F1 tells us
It is a tangible molecular candidate. No clear family match was found, which proves neither novelty nor function.
5. Structure is still a hypothesis
The available structure is a computational prediction. Further checks are needed before drawing any conclusion.
6. The CNB-1 evidence chain
A staged workflow, from generation to experimental validation, that every candidate must pass before biological claims can be made.
7. What comes next
Computational analyses for AGD-F1 and the PET-hydrolase candidate, then the generation loop again, across models and versions.
10. Conclusion
CNB-1 is producing testable molecular hypotheses, not a proven functional protein.