Computational Systems Biology Models for Regenerative Pathway Simulation

Abstract

Regenerative biology is governed by interlocking molecular signalling pathways -- Wnt, Notch, TGF-beta, Hedgehog, mTOR, and their downstream effectors -- whose dynamic interactions determine the balance between stem cell self-renewal and differentiation, tissue repair and fibrosis, and regenerative success and failure. Computational systems biology modelling of these pathways -- using ordinary differential equation (ODE) models, Boolean network models, and agent-based simulations -- provides a mechanistic framework for understanding pathway dynamics and predicting the effects of perturbations (genetic mutations, drug treatments, microenvironmental signals) on regenerative outcomes. This paper proposes the Regenerative Pathway Simulation Systems Biology (RPSSB) framework, a multi- formalism computational systems biology methodology for simulation of regenerative signalling pathways, comprising five modelling components: an ODE kinetic model builder (OKB) for quantitative signalling dynamics; a Boolean network regeneration simulator (BNRS) for qualitative pathway state analysis; a multi-scale agent-based tissue regeneration simulator (MABRS) integrating cell-level and tissue-level dynamics; a parameter estimation engine (PEE) fitting model parameters to experimental time-course data; and a perturbation response predictor (PRP) simulating the effect of genetic or pharmacological perturbations on regenerative pathway output. RPSSB is evaluated across four regenerative pathway systems: Wnt-Notch crosstalk in intestinal stem cell renewal, TGF-beta/Smad fibrosis-regeneration balance in liver, mTOR pathway regulation of satellite cell activation in skeletal muscle, and Hedgehog-mediated chondrocyte differentiation in cartilage. OKB reproduces experimental time-course data with mean R2 = 0.924 (SD = 0.028) across four pathway systems. BNRS identifies 84.6% of known stable regenerative and fibrotic attractor states. PRP predicts drug perturbation effects with AUC = 0.918 (SD = 0.022). MABRS simulations reproduce tissue-level regeneration dynamics with r = 0.884 (SD = 0.028) vs. experimental histological time-course data. The study contributes the RPSSB specification, four calibrated regenerative pathway models, and a ranked catalogue of 128 computationally predicted therapeutic perturbation strategies.