Digital Twins of Cell Differentiation Integrating Centrosomal Dynamics

Abstract

The integration of centrosomal dynamics into digital twins (DTs) of cell differentiation represents a paradigm shift from descriptive, empirical protocols to predictive, mechanism-based engineering of cell fate. Digital twins in biomedicine act as dynamic, multi-scale virtual systems that integrate experimental data and generate testable predictions, yet most models overlook the centrosome—an organelle increasingly recognized as an active signal-integration hub rather than a passive microtubule-organizing center. This review synthesizes evidence that centrosomes orchestrate cell fate decisions through asymmetric inheritance, CAFD (cell fate determinant) docking, primary ciliogenesis, and coordination with polarity complexes. We propose a six-layer digital twin architecture that incorporates the centrosome as a mandatory computational module at the perception–integration level and CAFD logistics at the intracellular transport level. Data acquisition analysis shows that super-resolution live-cell imaging, APEX2 proximity proteomics, FRET-based biosensors, and imaging mass cytometry provide multiparametric data to parameterize centrosome-specific dynamics. Computational frameworks include agent-based modeling with explicit centrosomal state vectors, hybrid LSTM–CNN architectures for multimodal time-series integration, graph neural networks for CAFD–centrosome interactomes, and reinforcement learning for closed-loop protocol optimization. A case study optimizing cortical neuron differentiation from iPSCs via Aurora A kinase inhibition illustrates implementation from data acquisition through hybrid model training. Key challenges include dimensionality, computational demands, incomplete CAFD inventories, and validation costs. Ethical issues of data sovereignty, liability, and reproducibility remain. Over three to five years, protocol twins will enable pharmaceutical in silico screening; within five to ten years, cloud platforms with pretrained centrosomal modules will democratize access; beyond ten years, integrated virtual cells may support Human Digital Twin initiatives. This interdisciplinary effort will accelerate predictable, safe cell therapies by transforming trial-and-error into centrosome-informed engineering.