[Depreciated and replaced by V3] From One Self-Proven Theorem to Blind Protein Structure: Blind Predictive Super Parity across 24 Sealed Whole-Structure Tests
Description
[Depreciated and replaced by V3] The application-specific clean rebuild has not yet been published; its authoritative theoretical boundary is now the governing V3 branch: From Fold to Life: An Exact, Zero-Parameter and Machine-Closed Foundational Reconstruction of Biology and Life Sciences from Smithian Fold Theory. The V3 source platform is https://github.com/MettaMazza/ernos-labs-sft-platform. The original DOI, concept DOI, version number and files are preserved for transparent historical provenance; this record must not be presented or cited as current V3 work.
Fold Protein achieves Blind Predictive Super Parity across 24 complete sealed whole-structure tests. Four preregistered panels generated and hash-sealed every state path and PDB before the corresponding experimental coordinates were accessed. Post-seal measurement records 0.9255486262 median TM_repo, 0.7833590149 Å median Cα RMSD95, and 0.9882113352 best whole-structure TM_repo. Fifteen of 24 predictions equal or outperform the 0.96 Å median Cα RMSD95 reported for AlphaFold at CASP14. Every execution records zero target accesses before sealing, zero trained weights and zero fitted parameters.
Super Parity is the combined evidentiary result, not a renamed prediction threshold. Fold Protein joins AlphaFold-class median backbone accuracy to a transparent first-principles derivation, exact finite state domains, machine-checked provenance back to the One, inspectable residue-by-residue execution, target-isolated sealing and public reproduction. An opaque trained predictor can establish reliability by repeatedly returning accurate answers; it does not expose or machine-prove the physical law responsible for those answers. Fold Protein predicts and supplies the explicit law it tests.
Protein Material Architecture V1 restores all 576 exact dihedral states at every residue, giving 43,776 residue-state trials. It applies the Smithian Fold Theory three-residue colour window, two-residue binary overlap and One-residue advance across generated material frames. The original complete 76-residue execution measures 0.9891211351 TM_repo, 0.2608575408 Å Cα dRMSD, and 0.3261459535 Å Kabsch Cα RMSD. Four subsequent panels preserve the frozen material relation and its sequence-only One-extension command across 24 independently deposited experimental structures.
The programme begins from Smithian Fold Theory's one machine-checked, self-proven theorem—there is no nothing—with zero axioms. The exact-fraction engine independently forces the canonical right-handed alpha-helix coordinates (−60°, −45°) and beta-sheet coordinates (−120°, +135°) on the complete signed 24-lattice. The complete evidence bundle preserves all favourable and unfavourable rows, registrations, seals, structures, hashes and executable comparison code.
The resource contrast is direct: this work was produced by one researcher on one Mac Studio in approximately one week divided among four SFT computational-proof programmes, against an AlphaFold comparator produced within Google's industrial research organization. Predictive parity is achieved; transparent derivation and proof establish Super Parity.
This standalone paper is the authoritative current synthesis of Fold Protein and closes the present study at repeated blind whole-structure predictive parity. The two earlier Protein papers remain preserved as superseded chronological development artifacts.
Open evidence and source: Fold Protein and Smithian Fold Theory of Everything.
Notes
Files
00_MAIN_PAPER_From_One_Self_Proven_Theorem_to_Blind_Protein_Structure.pdf
Files
(6.2 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:7afe61bd3e109ec6fbd1f1a0cb1ca3dc
|
153.3 kB | Preview Download |
|
md5:1491b1d5cd92c7fd254014740533063e
|
57.2 kB | Preview Download |
|
md5:1db0c21f714452164b9821950e9960cd
|
39.8 kB | Preview Download |
|
md5:9ab1d1754791ef78bb3ef4812f388b01
|
3.7 kB | Preview Download |
|
md5:b4e6e9ba75fd695612d1d7a00a0e53e3
|
20.6 kB | Preview Download |
|
md5:0994b482794b7a50e38522ab6baeba43
|
5.9 MB | Preview Download |
Additional details
References
- Levinthal, C. (1969). How to fold graciously. Mössbauer Spectroscopy in Biological Systems, 22–24.
- Parsons, J. et al. (2005). Practical conversion from torsion space to Cartesian space for in silico protein synthesis. Journal of Computational Chemistry, 26, 1063–1068.
- Zhang, Y. & Skolnick, J. (2004). Scoring function for automated assessment of protein structure template quality. Proteins, 57, 702–710.
- Jumper, J. et al. (2021). Highly accurate protein structure prediction with AlphaFold. Nature, 596, 583–589.
- Smith, M. (2026). The Smithian Fold Theory of Everything. doi:10.5281/zenodo.21182468.