Research into the biophysical properties of deoxyribonucleic acid (DNA) and the mechanisms underlying genetic mutations has undergone marked advancements. The intriguing nature of mutations resulting from retroviral DNA insertion has garnered considerable attention. Whether these mutations are random or region-specific, the distribution patterns of mutation sites have been the focus of numerous research endeavours. This mutation mechanism originates from interactions between host DNA and the pre-integration complex (PIC), comprising retroviral DNA and an integrase enzyme that facilitates its incorporation into the host DNA. Our study focused on the Zfp521 gene locus, recognised for its pronounced susceptibility to insertional mutations, particularly around unique palindromic sequences. We employed two biophysical models to predict mutation distribution within a range of 50 base pairs centred on these sequences. The first is a probabilistic collision model emphasising PIC and target DNA interactions. The second model is a DNA diffraction lattice, where the PIC behaves according to probability density. Although both models adeptly illuminated the probability distributions of target sites, the second model was more successful in predicting the PIC integration sites based on DNA biophysical properties. This highlights the pivotal role of intricate interactions between the PIC and target DNA, suggesting that mutations can be predicted in a stochastic manner.