Bayesian inference of agent-based models: a tool for studying kidney branching morphogenesis.

Area of Science:

• Developmental Biology
• Computational Biology
• Renal Physiology

Background:

• The adult mammalian kidney collecting duct epithelium forms via ureteric bud branching from the nephric duct.
• Branching morphogenesis is regulated by interactions between the ureteric bud epithelium and surrounding mesenchymal cells.
• Glial cell-derived neurotrophic factor (GDNF) signaling via RET receptors is crucial for ureteric bud branching.

Purpose of the Study:

• To develop the first agent-based model of ex vivo kidney ureteric branching.
• To investigate the role of growth factor-regulated mechanisms in epithelial cell branching.
• To infer model parameters and understand the dependency of growth factor signaling in kidney development.

Main Methods:

• Agent-based modeling of kidney ureteric branching.
• Comparison of model simulations with experimental data.
• Application of Approximate Bayesian Computation (ABC) for parameter inference.

Main Results:

• Growth factor-dependent mechanisms explain early epithelial cell branching.
• Epithelial cell division requires a switch-like response to local growth factor concentration, exceeding a threshold.
• Parameter inference revealed dependencies in the growth factor-dependent growth switch.

Conclusions:

• A threshold-dependent growth mechanism is critical for kidney collecting duct branching.
• GDNF signaling likely controls both proliferation and chemotaxis, essential for kidney morphogenesis.
• The model provides insights into the quantitative requirements for kidney development.