Skip to main content

​
Why Use This Engine?

In the documentation below, we will use Revilico’s Migration and Invasion assay engine to predict the effect of drug treatment on cancer cell motility and invasive capacity across cell lines with different inherent metastatic potentials. This assay is directly relevant to anti-metastatic drug development: a compound that kills cells in primary culture may or may not inhibit the migratory behavior that drives metastatic dissemination, and these two properties require independent evaluation.

​
Background

Cancer metastasis requires tumor cells to detach from the primary tumor, migrate through the extracellular matrix, invade through basement membranes, and establish colonies at distant sites. Cell migration is measured in vitro using two complementary formats: the scratch (wound healing) assay measures how quickly cells close a mechanically created gap in a cell monolayer, and the Transwell invasion assay measures how many cells actively migrate through a porous membrane coated with Matrigel (a reconstituted basement membrane extract) toward a chemoattractant. Cell lines differ substantially in their baseline migratory capacity based on their epithelial-mesenchymal transition (EMT) status. MDA-MB-231 (triple-negative breast cancer) is highly mesenchymal with a high basal migration rate. MCF7 (luminal breast cancer) retains an epithelial phenotype and migrates slowly. RevAssay models both assay formats using cell-line-specific basal migration rates and the GNN+MLP viability signal as the drug-induced suppression factor.

​
Simulation Model

Basal migration rates by cell line (calibrated to experimental scratch assay literature values): The drug-suppressed migration rate scales proportionally with predicted viability: rate(c)=R0⋅v(c)⋅η,η∼LN(0,0.152)\text{rate}(c) = R_0 \cdot v(c) \cdot \eta, \quad \eta \sim \mathcal{LN}(0, 0.15^2) Where R0R_0 is the cell-line basal rate and v(c)v(c) is the GNN+MLP viability at concentration cc. Scratch assay wound closure: Closure%(c)=v(c)×85+ε,ε∼N(0,82)\text{Closure\%}(c) = v(c) \times 85 + \varepsilon, \quad \varepsilon \sim \mathcal{N}(0, 8^2) Transwell invasion index (Matrigel-coated membrane): Invasion index(c)=v(c)×0.78⋅ηinv,ηinv∼LN(0,0.122)\text{Invasion index}(c) = v(c) \times 0.78 \cdot \eta_{\text{inv}}, \quad \eta_{\text{inv}} \sim \mathcal{LN}(0, 0.12^2) The invasion index is lower than wound closure because Matrigel provides an additional physical barrier that requires active matrix metalloproteinase (MMP) activity to penetrate. Kinetic time series for scratch assay wound closure over the time course TT: Closure%(t)=v(c)×90×(1−e−3t/T)+ε(t)\text{Closure\%}(t) = v(c) \times 90 \times \left(1 - e^{-3t/T}\right) + \varepsilon(t) This exponential approach to the asymptote reflects the decelerating rate of gap closure as the wound narrows.

​
Parameters

​
Outputs

  • Migration rate (um/h): Drug-suppressed migration rate per cell line per concentration
  • Wound closure %: Percentage of scratch gap closed at the assay endpoint
  • Invasion index: Normalized invasive capacity (Transwell mode only)
  • Kinetic curves: Time-series wound closure from 0 to the full time course
  • Dose-response curves: Migration rate and wound closure across the concentration range
  • Comparative bar chart: Cell lines ranked by migration rate, colored by drug concentration
  • 96-well heatmap: Plate-view colored by migration rate or wound closure fraction