Identifiability of phenotypic adaptation from low-cell-count experiments and a stochastic model

Area of Science:

• Quantitative biology
• Cancer research
• Mathematical modeling

Background:

• Phenotypic plasticity is a key factor in cancer treatment failure.
• Existing quantitative tools struggle to characterize plastic adaptation and distinguish resistance phenotypes.
• Low-cell-count proliferation assays are common experimental models.

Purpose of the Study:

• To develop a quantitative framework for analyzing phenotypic plasticity in cancer.
• To distinguish between discrete and continuous distributions of resistance phenotypes.
• To assess the identifiability of model parameters in proliferation assays.

Main Methods:

• Developed a stochastic individual-based model for plastic phenotype adaptation.
• Modeled a continuously-structured phenotype space in low-cell-count assays.
• Formulated a likelihood function capturing experimental noise.
• Applied the framework to assess parameter identifiability using population-level data.

Main Results:

• Cell-to-cell heterogeneity is practically non-identifiable from common experimental data.
• Population-level data are insufficient to differentiate discrete from continuous resistance phenotypes.
• Homogeneous ordinary differential equation models may adequately describe population behaviors.

Conclusions:

• Current population-level data and analysis methods are limited in characterizing cancer phenotypic plasticity.
• Future experimental designs and quantitative analyses need refinement to probe heterogeneity and resistance mechanisms.
• The developed model provides a framework for more robust analysis of adaptive resistance in cancer.