Evolutionary trajectory estimation via replication simulation of coronavirus spike gene based on random mutation and similarity-based selection.

Viruses exhibit rapid evolutionary dynamics through random mutations and selection, driving their adaptation and cross-species transmission. To investigate these mechanisms, we designed a simulation framework with a graphical user interface (GUI), implementing random mutation and similarity-based selection. This system models the evolution of a user-supplied viral sequence toward a designated target by recursively selecting the top-N amino acid sequences with the greatest similarity in each replication cycle. Simulations tracking the evolution of SARS-CoV-2 Wuhan-Hu-1 toward the Omicron variant (BA.1) displayed plateau-like similarity trajectories, where increased substitution rates resulted in a more rapid attainment of the plateau stage. The model-generated intermediate spike sequences exhibited similarities to real-world evolutionary patterns, including B, B.1.2, B.1.160, B.1.398, B.1.1.529, and BA.1 lineages. Additionally, the approach replicated the divergent evolutionary outcomes of PEDV subjected to distinct selection regimes (with and without trypsin treatment). While the model is simplified, it provides a means to explore plausible viral evolutionary paths and may contribute to identifying potential intermediates relevant to zoonotic spillover. Integrating features such as recombination, population-level effects, and further biological constraints could substantially enhance its predictive power in future iterations.