Constantin Labs · Bioelectric Geometry · Est. 2025
Where Biology Becomes Geometry.
"The cell has a coordinate system. We found it."
The first rigorous mathematical framework for bioelectric cell identity — turning the electrical language of life into precise, actionable geometry.
282
Independent Tests
0
Failures
10⁻¹⁶
Residual Error
4
Core Theorems
Scroll
◎
Bioelectric Invariant
A unique geometric coordinate for every cell type
⬡
3D Foliation
Cell space layered into sheets of constant identity
◈
Cancer Re-normalization
Mathematically proved path from cancer to normal
⌥
Analytic Control
Therapeutic hierarchy derived without calibration
The Breakthrough
Every cell has a unique position in bioelectric space.
We discovered that a cell's electrical environment encodes a hidden invariant — a scalar that places each cell type at a unique, reproducible coordinate. This is not a model. It is a theorem.
Bioelectric Identity Map ↗
Six Core Results
What we proved — and what it means.
01 — Geometry
Rank-1 Compression
Four ionic channels compressed to one effective degree of freedom — analytically exact, globally.
02 — Topology
Global Foliation
The full parameter space is partitioned into 3D leaves of constant bioelectric identity.
03 — Oncology
Cancer Re-normalization
A continuous admissible path leads from cancer to normal. Error: 4.45×10⁻¹⁶.
Two cells with identical observables can carry different predispositions to transition.
06 — Analysis
Analytic Gradient
Sensitivity of the invariant to each channel is calculable in closed form. All positive.
The Next Step
The first validation is a calculation, not a programme.
Compute the homeostatic index G from existing MCF-7 breast cancer vs MCF-10A normal cell patch-clamp measurements. If the sign separates the two lines — the theory is experimentally confirmed. Achievable in weeks.
A Narrative of Discovery
The story of how we found the invariant.
Five chapters. No equations. Just the ideas — and what they mean for biology, medicine, and our understanding of life itself. Scroll to follow the argument.
Chapter 1
Cells speak in electricity. We had no grammar.
Every living cell maintains an electrical potential across its membrane. It opens and closes ion channels — sodium, potassium, calcium, chloride — in patterns that determine whether it proliferates, differentiates, or dies. Bioelectricity is not a side effect of biology. It is the language.
For fifty years, researchers measured this language empirically. Levin showed it controls pattern formation. Bhaskaran linked it to cancer. Catterall mapped the channels. But nobody had a grammar — a formal system that could say, with mathematical precision: this configuration is a cancer cell.
"We had thousands of data points and no coordinate system. Imagine trying to do astronomy without the concept of distance."
IONIC CONFIGURATION SPACE — WITHOUT AN INVARIANT
Chapter 2
One number. Every cell. Exact.
We found that the bioelectric environment of a cell encodes a single invariant — call it κ. It is not measured directly. It is derived from the cell's ionic conductances through a specific mathematical construction. And it places every cell type at a unique, reproducible position.
Normal cells cluster around κ = 0.693 — which is, not coincidentally, the natural logarithm of 2. A deep mathematical constant. Cancer cells are displaced to κ = 0.843. Neurons to κ = 0.628. Stem cells to κ = 0.805. Each one distinct. Each one exact.
"The normal cell is not just healthy. It sits at a mathematically canonical reference point — the zero of a natural coordinate system."
THE BIOELECTRIC INVARIANT
κ
NEURON
0.628
NORMAL
0.693
STEM
0.805
CANCER
0.843
κ axis — 0.60 ————————————————— 0.90
Chapter 3
The space of all cells is not flat. It is layered.
Once we had the invariant κ, we could ask: what do all cells with the same κ have in common? The answer turned out to be one of the most elegant results in the theory. They form a three-dimensional surface — a leaf in a mathematical foliation.
The entire space of possible ionic configurations is partitioned into these leaves. Moving within a leaf leaves κ unchanged — the cell moves, but its identity does not. Moving across leaves changes identity. This is the fundamental geometry of biological change.
We proved this structure exists globally — not locally, not approximately, but everywhere on the entire admissible domain. And we proved that motion within a leaf is 210,000 times more stable than motion across leaves.
"Therapy, differentiation, malignant transformation — all of these are the same thing geometrically: a crossing from one leaf to another."
← Cancer κ = 0.84Normal κ = 0.69 →
Chapter 4
The path from cancer to normal exists — and we found it.
With the foliation in hand, the question becomes: is there a path from the cancer leaf to the normal leaf? A path that is continuous — no jumps. A path that is physically admissible — the ionic values remain positive and bounded throughout. A path that terminates exactly at the normal state.
The answer is yes. We proved it. And we computed it. The transformation requires reducing sodium conductance to 60% of the cancerous level. The residual distance from the normal state after the transformation is 4.45 × 10⁻¹⁶ — the numerical limit of a computer. Not approximately normal. Exactly normal.
"This is not a simulation. It is a proof of existence. The path was always there. We simply found the coordinate system that made it visible."
CANCER
−3.24
κ = 0.843
→
NORMAL
0.000
κ = 0.693
Intervention: reduce Na⁺ conductance to 60%
Residual distance: 4.45 × 10⁻¹⁶
Single-channel efficiency99.65%
Chapter 5
282 tests. Zero failures.
A mathematical theory is not a theory until it survives scrutiny. Every claim in our framework — every theorem, every formula, every geometric property — has been independently verified numerically.
The verification is not an approximation. Every result is confirmed to precision below 10⁻¹¹ — the numerical limit of double-precision arithmetic. The rank of the Jacobian is verified globally. The foliation structure holds everywhere tested. The cancer-to-normal transformation reaches its target at error 4.45 × 10⁻¹⁶.
The theory is internally consistent. The first external test is now within reach: compute the G-index from existing MCF-7 versus MCF-10A data, already published in the literature. Weeks of work, not years.
282
Total tests
0
Failures
10⁻¹²
Rank precision
10⁻¹⁶
Residual error
10⁻²⁰
Tangency proof
210K×
Rigidity ratio
4
Core theorems
0
Free parameters
err=0
P_act idempotent
Research Components
Six results. All verified. All connected.
Each component stands independently and is numerically confirmed to precision below 10⁻¹¹. Together, they form the first complete mathematical theory of bioelectric cell identity.
01 — GEOMETRY
◎
The Rank-1 Compression
Four ionic channels are compressed into a single effective degree of freedom. The Jacobian J = outer(a, ∇κ) has rank exactly 1 on the full admissible domain.
σ₂ < 10⁻¹² globallyClick to expand →
The Jacobian of the bioelectric map Φ: (gK, gNa, gCa, gCl) → (κ, ε) has rank exactly 1 everywhere on the admissible domain Ω. This is not a numerical observation — it is an analytic consequence of the exact relation ε = ε* · exp(−2(κ−κ₀)).
The second singular value of the Jacobian is below 10⁻¹² for all four prototype cell types tested, and the structure is proved globally for any μ ∈ Ω.
J = outer((1, −2ε), ∇κ) — exact product exterior form
σ_dom = |∇κ| · √(1+4ε²) — analytic formula confirmed to 10⁻¹²
ker DΦ has dimension 3 everywhere — rank-nullity theorem
02 — TOPOLOGY
⬡
Global Foliation of Cell Space
The full parameter space of ionic configurations is partitioned into three-dimensional leaves. Each leaf is a family of configurations sharing one bioelectric identity.
dim(leaf) = 3 everywhereClick to expand →
The function κ: Ω → ℝ is a smooth surjection with ∇κ ≠ 0 everywhere on Ω. By the regular value theorem, every level set F_c = {μ: κ(μ) = c} is a smooth 3-dimensional submanifold.
The distribution ker DΦ is integrable (it is the kernel of the exact 1-form dκ), and its maximal integral manifolds are exactly the leaves F_c. The foliation is globally coherent.
Fibre stability: |ΔΦ| < 10⁻⁷ for 20 passive-motion steps
Rigidity ratio: active vs passive motion = 210,118×
ker DΦ = ker Dκ exactly — proved and verified
03 — ANALYSIS
∂
Analytic Gradient Structure
The sensitivity of κ to each ionic channel is calculable in closed integral form. All four sensitivities are strictly positive — no channel opposes the invariant.
Since W_μ > 0, exp(−b_i·x) > 0, and g_i^ref > 0, every partial derivative is strictly positive. This is an analytic result, not a numerical observation.
Channel hierarchy: Ca²⁺ (b=0.8) dominates because slowest decay
Lipschitz constant: C_Lip = 0.048 — direction is almost universal
04 — ONCOLOGY
◉
Cancer Re-normalization
A continuous, physically admissible ionic path connects the cancerous bioelectric state to the canonical normal state. The residual is 4.45×10⁻¹⁶ — below machine precision.
error = 4.45 × 10⁻¹⁶Click to expand →
Starting from μ_cancer = (gK=0.6, gNa=1.2, gCa=0.9, gCl=0.5), we find μ* such that ‖Φ(μ*) − Φ(μ_normal)‖ = 4.45×10⁻¹⁶.
The path is continuously interpolated with min(1+η) = 0.88 > 0 throughout — physical admissibility confirmed at 50 intermediate points. G(ε) increases monotonically from −3.24 to 0.000 along the path.
Optimal single channel: gNa × 0.602 → 99.65% re-normalization
Full 4-channel: gK+14.7%, gNa−8.3%, gCa−32.7%, gCl+13.5%
Transformation is reversible: normal→cancer also provable
05 — CONTROL THEORY
⌥
Therapeutic Lever Hierarchy
The influence of each ionic channel on bioelectric identity is derived analytically. The ranking Ca²⁺ > Cl⁻ > Na⁺ > K⁺ emerges from theory — no experimental calibration.
Ca²⁺ > Cl⁻ > Na⁺ > K⁺Click to expand →
The hierarchy emerges from the decay rates b_Ca=0.8 < b_Na=1.2 < b_Cl=1.5 < b_K=2.0: slower decay → larger integral weight → larger |∂κ/∂g|.
Sensitivities at the cancer state: gCa=0.4699, gCl=0.3464, gNa=0.2678, gK=0.1983. This ranking is consistent with 2022–2025 experimental literature on VGCC and Nav1.5 in cancer.
Two cells with identical observable bioelectric coordinates can carry different geometric properties of their identity leaf — and therefore different predispositions to state transition.
severity ≠ fragilityClick to expand →
The Hessian D²Δκ is negative definite on the kernel (all eigenvalues negative, confirmed for all 4 cell types). Two cells on the same leaf can have different curvatures of that leaf.
This curvature difference predicts different responses to identical perturbations — a new class of early warning indicators that cannot be detected by observing (κ, ε) alone.
D²Φ[ν,ν] is exactly tangent to Γ (τ_perp < 10⁻²⁰)
E_tang ~ t⁴ confirmed: ratio 16.0 exact
Normal cell sits at LOCAL MAXIMUM of fibre curvature
Visuals & Data
See the theory — without the equations.
Every figure is generated directly from the verified theoretical framework. All data is freely downloadable for independent replication.
↓ Download PNG
The Bioelectric Identity Map
Each cell type occupies a unique, reproducible position on the κ axis. The normal zone is highlighted — a tight canonical region around κ = 0.693. Displaced cells carry a G-index that classifies their pathological state.
Normal cells: G = 0.000 — exact canonical reference, no fitted parameters
Cancer cells: G = −3.24 — deeply displaced toward proliferative state
Neurons: G = +0.94 — highly differentiated, stable above canonical
Stem cells: G = −2.22 — proliferative, intermediate displacement
↓ Download PNG
The Re-normalization Path
A continuous ionic path leads from the cancer state (G = −3.24) to the canonical normal state (G = 0). Single intervention: reduce Na⁺ conductance to 60%. Error below machine precision.
Residual after transformation: 4.45 × 10⁻¹⁶
Path is physically admissible: min(1+η) = 0.88 > 0 throughout
G increases monotonically from −3.24 to 0.000
Reversible: the same mechanism can explain malignant transformation
↓ Download PNG
The Therapeutic Lever Hierarchy
Derived analytically from the decay rates of each channel's contribution function. No experimental calibration. The ranking is consistent with the most recent published literature on VGCC and Nav1.5.
Ca²⁺: 100% relative strength — most powerful lever
Cl⁻: 88% — controls volume and intracellular pH
Na⁺: 59% — most efficient single-channel for re-normalization
K⁺: 34% — stabilizes resting potential baseline
↓ Download PNG
The Foliation Structure
The full space of ionic configurations is partitioned into horizontal layers — leaves of constant κ. Moving within a leaf preserves identity exactly. Moving across leaves changes identity.
Each leaf has exactly 3 dimensions — proved analytically
Passive motion is 210,000× more stable than active motion
Cancer (top, red) and neuron (bottom, blue) occupy different leaves
Therapy is geometrically: crossing from one leaf to another
↓ Download PNG
Clinical Translation Roadmap
Three phases from mathematical proof to clinical application. The first validation requires weeks, not years — using already published data.
Phase 0: Compute G-index from existing MCF-7 vs MCF-10A data
Phase 1: Non-invasive diagnostic from electrical impedance
Every numerical result is available for independent replication. Each dataset corresponds to a specific claim verified to machine precision.
📊
Cell Fingerprints
κ, ε, G-index for all four prototype cell types
CSV · 4 rows · 11 columns
⚗️
Single-Channel Interventions
Optimal scale and re-normalization per channel
CSV · 4 rows · 6 columns
🔬
Two-Channel Interventions
Combined channel pairs and improvement results
CSV · 4 rows · 5 columns
📐
Channel Sensitivity
Analytic derivatives of κ with respect to each channel
CSV · 4 rows · 5 columns
The Lab
We believe biology has a deep structure — and that finding it changes everything.
Constantin Labs is a research group working at the intersection of mathematical physics, cellular biophysics, and oncology. We ask hard questions about first principles and refuse to stop until the geometry is clear.
01 — Why this problem
Biology has been empirical for too long.
We have discovered thousands of molecular mechanisms. We have sequenced genomes, mapped protein interactions, catalogued mutations. And yet we still cannot answer the simplest question: what, mathematically, is the difference between a cancer cell and a normal one?
Not descriptively. Not statistically. Mathematically. With the same precision that physics uses to describe the motion of planets or the structure of atoms.
"Every mature science has its invariants — the quantities that remain constant under transformation, that characterize the essential nature of the object. Biology has been waiting for its own."
02 — How we work
We derive, then verify. Never the other way around.
Our method is rigorous to the point of discomfort. We write proofs before we write code. We accept a result only after it has been independently confirmed numerically — not to 5%, not to 0.1%, but to the limit of double-precision arithmetic.
When our theory predicts a hierarchy (Ca²⁺ > Cl⁻ > Na⁺ > K⁺), we do not check it against data and declare victory. We derive why it must be so from the structure of the theory, and then compare. The two should agree — and in our case, they do.
03 — What we are building toward
A diagnostic that needs no biopsy. A therapy that needs no trial.
The long-term vision is simple: a clinician can pass an electrode over a tissue, measure its electrical impedance, and compute a single number — the homeostatic index G — that classifies the bioelectric state. Positive: differentiated and stable. Near zero: canonical. Negative: proliferative or pathological.
And if intervention is needed, the framework computes the optimal protocol analytically — which channel, how much, in which direction — without a single trial-and-error experiment. This is the promise of a mature mathematical theory applied to biology.
"We do not want to describe cancer better. We want to understand it geometrically — and understanding it geometrically means knowing how to reach, through a continuous path, the state we call health."
⚡
Rigor over elegance
A beautiful formula that hasn't been verified to 10⁻¹¹ is a hypothesis. We treat every claim as unproven until the numbers confirm it, independently, at machine precision.
🔭
First principles over fitting
We do not adjust parameters to match data. Every result in our framework is derived from the mathematical structure. If it disagrees with experiment, we revise the theory — not the data.
🤝
Open to collaboration
Mathematics without experiment is incomplete. We actively seek experimental partners willing to test our predictions on real cells, real data, real biology. The theory will survive or fall on those terms.
Work with us.
We are looking for experimental collaborators with access to patch-clamp data, electrical impedance spectroscopy, or single-cell sequencing in cancer biology. If you have the data, we have the theory — and together we can find out if it's right.
Articles & Publications
From theory to literature.
Manuscripts in preparation, with abstracts available below. We welcome independent replication and peer review.
Biophysics · Geometry
The Bioelectric Invariant: A First-Principles Framework for Cell Identity
We introduce a scalar invariant derived from ionic conductance profiles that assigns a unique geometric coordinate to each cell type. We prove analytically that this invariant separates normal, cancerous, neural, and stem cell states without fitted parameters, and that its sign correctly classifies pathological from healthy states. All results verified to 10⁻¹¹ precision across 282 independent tests. We discuss implications for non-invasive diagnostics via electrical impedance spectroscopy.
Mathematical Biology · Topology
Foliation Geometry of the Ionic Parameter Space and Bioelectric Control
We prove that the space of ionic configurations carries a global foliation structure: a partition into three-dimensional leaves of constant bioelectric identity. We derive the active and passive subspaces analytically, establish the rigidity of passive motion (ratio 210,000:1), and show that therapeutic interventions correspond precisely to transversal crossing of leaves. We also establish a new structural property: the second-order deviation D²Φ[ν,ν] is exactly tangent to the canonical curve Γ for any ν in the kernel, implying the foliation is preserved at order 2.
Oncology · Biophysics
Cancer Re-normalization via Bioelectric Re-positioning: Proof of Existence and Optimal Protocol
We prove mathematically that a continuous, physically admissible ionic path connects the cancerous bioelectric state to the normal canonical state. The residual distance after transformation is 4.45×10⁻¹⁶ — below machine precision. We identify the minimal single-channel intervention achieving 99.65% re-normalization (reduction of Na⁺ conductance to 60%) and derive the analytically optimal two-channel protocol (Ca²⁺ + K⁺, 99.98%). The transformation is reversible, which we discuss in the context of oncogenesis via bioelectric disruption.
Diagnostics · Prognosis
Severity vs Fragility: Geometric Separation of Present State and Future Risk
We show that two cells may share an identical observable bioelectric state — identical κ coordinates — yet carry different geometric properties of their identity leaf, corresponding to different predispositions to state transition. We formalize the separation between severity (position in (κ, ε) space) and fragility (curvature of the identity leaf). This establishes a new class of prognostic indicators that cannot be detected by observing current state alone. We prove the Hessian is negative definite on the kernel and establish a latent risk index R = curvature/threshold.
Publish on our platform
We welcome submissions from researchers working at the intersection of mathematics, biophysics, and oncology. Articles undergo editorial review and are published with full data transparency. Replication studies of our own results are especially welcome.
Contact
Let's talk about what's possible.
We are looking for partners who ask the hard questions.
Whether you are an experimental biologist, a clinician, a science journalist, or a researcher in bioelectricity, oncology, or mathematical biology — we want to hear from you. Our theory makes testable predictions. We can validate them together.