Precision Oncology Biomarker Matching

Targeted Therapy

Dose intensity 70%

Add checkpoint-inhibitor combo (anti-PD-L1)

Mutation rate (resistance drift) 1.5%

Tumor Status

Tumor burden

100%

Weeks elapsed

Selected-marker positive fraction

Cells killed this cycle

–

Low marker expression

High marker expression

How it works

Each cell in this synthetic tumor carries its own gene-expression profile — continuous levels of three biomarkers (HER2, EGFR, PD-L1) drawn from four founder subclones, so the mass starts genetically heterogeneous just like a real tumor:

Subclones (baseline HER2, EGFR, PD-L1):
  HER2-driven    (0.80, 0.15, 0.15)  ~35%
  EGFR-driven    (0.15, 0.80, 0.15)  ~30%
  PD-L1-high     (0.15, 0.15, 0.75)  ~20%
  Triple-low     (0.10, 0.10, 0.10)  ~15%  (intrinsically resistant)

A targeted therapy only kills cells whose matching marker is actionably expressed. Per weekly cycle, kill probability for a cell uses a smooth threshold (smoothstep between 0.3 and 0.9) on the chosen marker, scaled by dose:

kill(cell) = dose × smoothstep(0.3, 0.9, expr[target])
if combo:  kill(cell) = max(kill(cell), 0.7·dose × smoothstep(0.3,0.9, expr[PD-L1]))
cell dies this cycle with probability kill(cell)

Surviving cells repopulate toward capacity each cycle; each new cell inherits its parent's expression profile plus Gaussian mutation noise (the "mutation rate" slider). Because a single-target therapy leaves low-expressing subclones almost untouched, they are the ones left to repopulate — this is clonal selection, the same mechanism behind acquired resistance in real precision oncology. Watch the biomarker-positive fraction fall while the resistant triple-low population quietly takes over; a checkpoint-inhibitor combo broadens the kill and slows that drift.

Marker view: HER2

Low expression High expression

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.