Simulations of deoxyribonucleic acid (DNA) molecular damage use the traversal algorithm that has the disadvantages of being time-consuming, slowly converging, and requiring high-performance computer clusters. This work presents an improved version of the algorithm, “density-based spatial clustering of applications with noise” (DBSCAN), using a KD-tree approach to find neighbors of each point for calculating clustered DNA damage. The resulting algorithm considers the spatial distributions for sites of energy deposition and hydroxyl radical attack, yielding the statistical probability of (single and double) DNA strand breaks. This work achieves high accuracy and high speed at calculating clustered DNA damage that has been induced by proton treatment at the molecular level while running on an i7 quad-core CPU. The simulations focus on the indirect effect generated by hydroxyl radical attack on DNA. The obtained results are consistent with those of other published experiments and simulations. Due to the array of chemical processes triggered by proton treatment, it is possible to predict the effects that different track structures of various energy protons produce on eliciting direct and indirect damage of DNA.

脱氧核糖核酸（DNA）分子损伤的模拟使用遍历算法，该算法具有耗时，收敛缓慢且需要高性能计算机集群的缺点。这项工作使用KD提出了算法的改进版本，即“基于噪声的应用程序的基于密度的空间聚类”（DBSCAN）。-tree方法找到每个点的邻居，以计算簇状DNA损伤。所得算法考虑了能量沉积和羟基自由基攻击位点的空间分布，得出了（单双链）DNA链断裂的统计概率。在i7四核CPU上运行时，这项工作在计算分子水平上由于质子处理引起的簇状DNA损伤时，可以实现高精度和高速度的计算。模拟的重点是羟基自由基攻击DNA所产生的间接影响。所得结果与其他已发表的实验和模拟结果一致。由于质子处理引发的一系列化学过程，