Potency only counts where the drug arrives.

The problem

A potency value is a statement about a concentration: at this concentration, this fraction of cells responds. It says nothing about whether that concentration ever exists anywhere in a tumour.

So two compounds can carry the same potency and behave nothing alike in tissue, and the ranking that came out of the plate can be exactly inverted by the geometry. This is not an edge case. It is one of the ordinary ways a programme spends two years finding out something the first spatial calculation would have said.

What LURA computes

Exposure is where the transport field and the gradient fields are read together against the compound activity. For every position in the modelled tissue, at every point in the schedule, the question is simple: is the free concentration here above what this compound needs, in the local conditions here.

• The share of tumour volume that reaches a therapeutic concentration
• How long it stays there, because a brief crossing is not an exposure
• Where the shortfall is, described as a region rather than as a deficit
• How that changes with dose, with schedule, and between candidates

What comes out

The comparison a portfolio decision actually needs: not which molecule is most potent, but which one covers most of the tumour it is meant to treat, and what it would take to cover the rest.

Where the answer is that a region is never reached, LURA says so plainly and names it, instead of returning a lower average that hides it. A named failure is a design brief. An averaged one is a surprise later.