01
Why movement matters
A molecule must cross spaces before it can interact with its target. A strong binder is not useful where it cannot arrive.
Working Study / Spatial mechanisms 01
The molecule that binds perfectly but never arrives.
Before target engagement comes transport. Explore how movement, supply, clearance and local capture determine which cells can even encounter a molecule.
Design a counterfactual
Change the journey
The parent stays fixed as a control. Modify the candidate and watch how its free-material profile changes in exactly the same abstract space.
Modeled free-material profile
From vessel edge to distant cells
Parent Candidate
Mean free level
26.2%
Parent: 19.3%
Distant-edge level
5.2%
Parent: 1.0%
Space above 20% of supply
56.1%
Parent: 41.5%
What the experiment teaches
More of the candidate's free material reaches the far edge in this model.
This measures only free material in a toy transport model. Stronger local binding can reduce the free fraction at a distance; that does not prove lower target engagement or worse treatment response.
Numerical mass-balance residual: 6.11e-16 · fixed step / stable diffusion discretization
02
Why binding can mislead
Local capture can hold material near the vessel while leaving less free material deeper in tissue. Capture and useful engagement are not identical.
03
What would validate this
Measure spatial concentration over time, compartment geometry, tissue-specific transport and binding, then test against independent observations.
Spatial mechanisms 02 / The compartments
Arriving in tissue is not arriving at the target.
Follow a finite abstract dose from a vessel, into extracellular space, across a cell membrane, and finally into a reversible target-bound pool. Each barrier can change the journey without changing the target itself.
Modify the candidate
The parent is fixed. Change one assumption at a time and look for the compartment where the two journeys separate.
Where the material goes
Final fractions of the same starting amount; removed material is counted, not lost from the accounting.
Vessel
0.5%
Parent 1.2%
Outside cell
2.9%
Parent 8.1%
Inside cell
13.4%
Parent 11.3%
Bound
43.6%
Parent 35.6%
Removed
39.6%
Parent 43.7%
━ Parent━ Candidate
Integrated free-cell amount
1.538
Parent 1.201
Integrated bound amount
2.871
Parent 2.178
What this teaches
The candidate assumptions increase time-integrated target capture. Which transfer step caused it—and could a real molecule achieve that change?
Conservation residual: 2.22e-16
Mechanism analysis / Controlled comparisons
Which design change actually matters?
The complete candidate changes several things at once. Here we repeat the simulation five times, switching only one assumption from parent to candidate each time. This reveals which changes are individually influential within this specific model.
| Mechanism stage | Parent → candidate | Isolated change in integrated bound amount |
|---|---|---|
| Vessel exit | 0.80 → 0.80 | +0.0000 |
| Membrane entry | 0.65 → 1.15 | +0.5268 |
| Cellular efflux | 0.45 → 0.25 | +0.1952 |
| Target capture | 0.75 → 0.75 | +0.0000 |
| Target release | 0.25 → 0.25 | +0.0000 |
Complete candidate change
0.6929
Sum of isolated changes
0.7220
Non-additive remainder
-0.0291
Learning prompt
The largest isolated effect here comes from membrane entry. That is a useful hypothesis about the model, not proof that this property is achievable in a real molecule.
Changes can interact: the complete effect need not equal the sum of isolated effects. This remainder is a mathematical diagnostic, not evidence of a specific biological interaction.
What evidence would we need?
Membrane entry
A molecule can be abundant outside the cell but scarce inside it.
Permeability or cellular uptake with a matched extracellular concentration.
Cellular efflux
Export competes with intracellular residence and target contact.
Transporter expression, inhibitor controls, and measured efflux kinetics.
One-factor changes are controlled comparisons around the fixed parent, not independent additive causes. The interaction residual combines nonlinearity and joint parameter changes; it is not a separately identified biological mechanism. All rates and outputs are dimensionless educational assumptions, not measured drug properties or efficacy.
Scientific boundary: Educational finite-bolus four-pool transport model only. All pools have equal abstract volume; rates, capture and removal are dimensionless assumptions. Target-bound material is not equivalent to receptor occupancy, therapeutic action, selectivity, or clinical exposure. The experiment does not model real membranes, transporter proteins, receptor densities, physical volumes, metabolic pathways, tissue geometry, or chemical structure. These are questions for future calibration, not properties inferred from the sliders.