We build mathematical frameworks that describe how complex systems maintain stability. Our first product, Angelika Fold, applies Recursive Coherence Field Theory to protein stability prediction.
RCFT started with a question that wouldn't go away: why do we have dozens of metrics for protein stability, but no single framework that tells you why a protein is stable? Rosetta gives you an energy score. AlphaFold gives you a structure. FoldX gives you a ΔΔG. But none of them give you a unified picture of how all the forces — hydrophobic packing, electrostatics, hydrogen bonds, backbone strain, evolutionary pressure — interact to produce the thing we call "stability."
The insight behind RCFT is that stability isn't a number you measure — it's a field you compute. A protein is stable when its internal states are mutually consistent: when the hydrophobic core supports the electrostatic network, which supports the hydrogen bond pattern, which supports the backbone geometry, which supports the hydrophobic core. It's recursive. The stability of each part depends on the stability of every other part.
That recursion is the core of RCFT. The framework doesn't just score each property independently and add them up. It evaluates how well they agree with each other, iterating until it finds the fixed point — the Ω score — where the system's self-assessment converges.
The golden ratio φ shows up naturally in the convergence rate. We didn't put it there. It emerged from the mathematics of recursive self-evaluation. Make of that what you will.
We publish what we know and what we don't. Our benchmark shows a strong ranking signal but the classification threshold needs calibration. We say that upfront because trust is worth more than a press release.
RCFT is a mathematical framework, not a brand. The Ω metric either works or it doesn't, and the only way to find out is rigorous, reproducible testing against experimental data. We welcome scrutiny.
The core RCFT engine runs on private infrastructure. We don't expose proprietary algorithms in client-side code, and we don't collect user data. If you submit a protein sequence through our demo, it's processed and discarded.
We're not trying to replace AlphaFold or Rosetta. We're building something complementary — a stability coherence layer that could sit on top of existing structural prediction tools. If you're a lab with experimental stability data, we want to talk.
If you're a researcher with experimental protein stability data, a biotech company interested in licensing, or an investor who understands that real science takes time — we'd like to hear from you.
© 2025 ERIC L. MARX — MARX SYSTEMS LLC — ALL RIGHTS RESERVED
PATENT PENDING — RCFT & ANGELIKA FOLD