Inverse Modeling Reveals Governing Law of Epithelial Migration
Kikuchi et al. (published December 29, 2025 in PLoS Computational Biology) develop a machine-learning inverse-modeling framework that infers governing equations from live-cell imaging and apply it to MDCK epithelial sheet migration driven by MAPK/ERK. The learned equations accurately predict single-cell movements from local chemical and mechanical cues, reveal spatiotemporal derivative processing and cell-to-cell heterogeneity. They release code and datasets and refine forward models to improve tissue-scale predictive simulations.
Key Points
- 1Predicts single-cell displacements from local ERK activity and mechanical cues with high quantitative accuracy.
- 2Reveals cells compute spatial gradients and temporal derivatives, clarifying signal-processing mechanisms driving migration.
- 3Enables refinement of forward models and provides reproducible code for predictive tissue-scale simulation improvements.
Scoring Rationale
High methodological novelty and reproducible code, limited by focus on epithelial migration dataset and niche biological scope.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
