Modeling Radiation-Induced Chromosome Aberrations Using a Polymer Physics-Based Nucleus Model
Abstract
Purpose
Differences in biological effects among radiation qualities arise from variations in the spatial distribution of initial DNA damage and subsequent repair processes, both are closely linked to chromatin organizations. This study aimed to integrate recent advances in chromatin architecture into radiobiological modeling.
Methods
Chromatin interactions were simulated using a polymer physics framework combined with a multi-stage relaxation strategy to decouple structural levels. The model accounts for the biomechanical properties of chromatin fibers, including tension, repulsion, and bending stiffness, as well as structural features revealed by Hi-C experiments. A hierarchical search algorithm and an inverse transformation mechanism were implemented for efficient DNA damage scoring, and a distance-dependent non-homologous end joining model with graph theory-based connected component analysis was used to simulate chromosome aberrations. Experimental data of human skin fibroblasts irradiated by γ-rays and α particles were used for benchmarking.
Results
The proposed method can generate interphase nucleus models of arbitrary size, shape, and chromosome number. The multi-stage relaxation reduced computational time by over three orders of magnitude compared with conventional polymer relaxation approaches. For a human diploid cell, the entire process took approximately three hours. The model reproduced key chromatin features, including chromosome territories, subcompartments, and TADs, consistent with the general patterns revealed by Hi-C data. Predicted contact maps and probabilities aligned with experimental measurements. The multi-stage relaxation demonstrated strong robustness across a wide range of hyperparameter choices, including maximum step size, random motion, and repulsion range. Simulation of dicentrics, interstitial deletions, and total aberrations agreed with experimental observations within 20% for both γ-rays and α particles.
Conclusion
The proposed polymer physics-based model bridges structural and radiobiological simulations, offering accurate and efficient prediction of DNA damage and chromosome aberrations. This approach enables realistic whole-nucleus simulations on standard computing platforms, with potential applications in particle therapy and space radiation risk assessment.