Declare the measurement
Fix preparation, units, time horizon, future panel and a functionally meaningful equivalence margin. Verify the assay and matching error.
The next focused expedition
The biology
conservatory.
A living map, built one measured connection at a time. State, energy, control and the possibility of recovery.
Sub-lab 01 / Bounded reaction ledger
What actually limits an enzyme?
Start with the smallest living mechanism we can fully specify: a finite enzyme pool converting a finite substrate pool inside a declared volume.
The instrument keeps three descriptions side by side. A linear benchmark treats rate as indefinitely scalable. Michaelis–Menten adds saturation. The finite-pool ledger then asks the harder question: after a real time interval, how much substrate remains, and which ceiling became active?
This is a model comparison, not a measurement. Every dial is declared, every unit is visible, and the output tells us what experiment would separate the models.
Reference foundations: NCBI enzyme kinetics and the mass-action versus Michaelis–Menten review.
Interactive model / declared assumptions
Turn the ceilings.
Set enzyme concentration, substrate concentration, catalytic turnover, affinity and observation time. The finite ledger integrates the saturating rate only until the substrate pool is exhausted.
- Well-mixed fixed volume.
- Initial-rate comparison; no product inhibition or transport.
- Nonnegative enzyme and substrate pools.
- Michaelis–Menten is admitted only as a comparison law.
The linear benchmark uses the low-substrate tangent, v = VmaxS/Km. It is allowed to exceed Vmax so the missing ceiling is visible.
Enzyme ledger / illustrative units
Vmax 0.004MM rate 0.003Linear rate 0.008Finite product 0.080
Active constraint: enzyme capacity or time
Rates are mM/s after the concentration-unit conversion µM → mM. Product is capped by the available substrate. A result here is a calculation receipt, not an enzyme assay.
Our direction
When does “recovered”
really mean recovered?
Make this the first unifying biological question. A restored reading can conceal a depleted pool, a slow controller or a changed response to the next challenge.
The papers already supply the parts of a serious programme: finite-resource accounting, competing adaptive channels, history-conditioned state tests, a proposed rescue-window law and the distinction between prediction and action. The next achievement should connect those parts in one independently measured system.
Recommended lead system: redox regulation in a tractable cultured-cell preparation, with a separately validated chemical assay. The immediate work is assay and data readiness. No animal or human experiment is launched by this website.
The shared biological map
Follow what enters.
Measure what remains.
| Workstream | Observable and action | Connection it must earn |
|---|---|---|
| Resources | Absolute pools, supply, consumption and regeneration; calibrated perturbation. | A reserve estimate predicts something a fraction alone misses. |
| Regulation | Signal fidelity, controller response time, capacity and surveillance readout. | A selective intervention distinguishes a slow controller from a depleted resource. |
| History and state | Different prior exposures, matched present and the same future challenge. | A retained distinction improves prediction in independent units and batches. |
| Recoverability | Specified rescue, withdrawal, rechallenge and independently measured durable fate. | Apparent recovery predicts continued function after support ends. |
| Energy and metabolism | Supply/demand and coupling efficiency under independently estimated losses. | The predicted binding-versus-slack interaction survives a controlled comparison. |
| Cells and communities | Lineages, growth, division, labelled trajectories and exchanges. | Individual-state rules transfer only through measured interactions and observation rules. |
First campaign / proposed protocol
The recovered-present test.
Create different pasts
Use controlled histories that overlap at the chosen present measurement. Preserve same-history controls and all intervention records.
Challenge the present
Apply the identical future. Compare a frozen snapshot baseline, a conventional mechanistic model, and the predeclared resource/history extension.
Test the repair
Freeze the added representation. Evaluate new histories and independent batches. Demand both better prediction and a controlled residual-history test.
What counts as success?
The initial measurement shows a resolved, functionally relevant common-future difference beyond matching error. An independently measurable added quantity improves a predefined held-out prediction score over strong baselines, and its residual-history discrepancy is inside the predefined equivalence margin.
What counts as failure?
Adequate equivalence supports the simpler representation in the tested scope. No gain rejects the proposed addition. Poor overlap, noisy assays or a confidence interval crossing the decision boundary give unresolved, not success.
Before substantial experimental work
Choose the biological endpoint and pilot-estimated precision, establish data access, then freeze effect margins, sample-size justification, split rules and stopping criteria. The website does not invent numerical success thresholds before those measurements exist.
A bounded-resource lens
A fraction cannot
tell you the reserve.
Hold the reduced fraction fixed while changing the total glutathione pool. The available accounting budget changes even though the fraction is identical.
In the declared fixed-volume GPx/GR ledger, GSH/2 + NADPH is a necessary resource budget for peroxide-equivalent processing. It assumes nonnegative pools and the specified stoichiometry; it does not establish a reaction rate, reachable endpoint or survival.
Synthetic resource ledger / mM
GSH 1.80Necessary budget 1.10
GSH = total glutathione equivalents × reduced fraction. Budget = GSH/2 + NADPH. No external supply, no rate or viability prediction. This is an explanatory model, not a biological measurement.
What the Observatory already taught us
A useful signal.
And an important rejection.
| Case | Result already recorded | Meaning for the new programme |
|---|---|---|
| Chlamydomonas model | Equal total light over 48 hours produced one cell versus eight descendants under different schedules, with nearly equal total lineage volume. | A synthetic example of timing and readout separation in a published model. It does not establish a new algal mechanism. |
| Numerical stress test | An apparent hidden-state lag vanished under a more accurate solver. | Numerical convergence is part of the evidence boundary. |
| Held-out cell prediction | Adding growth-history data worsened overall Brier score by 3.30% on the recorded 3,679 predictions. | More history is not automatically useful. The proposed extra state must earn its place. |
IDA / A distinct research workstream
How does a system
find its way back?
IDA asks whether the path of return after perturbation contains information that an instantaneous reading misses. Its first gate is a real-signal prediction test against static and conventional time-series baselines, with person and session holdouts.
Only after that gate should feedback be tested against matched replay, sham and conventional control. Gamma-to-belief and inter-brain sinh coupling are separate hypotheses. None of these results would, by itself, explain consciousness.