Exploring Protein Knots with Artificial Amino Acids: Insights from AlphaFold-Driven Molecular Simulations.

Knots in proteins are entangled topological features, influencing protein stability and function. However, the mechanisms behind protein knotting and maintenance are not widely understood, limiting the ability to improve protein engineering design. We aim to investigate how modifications to knotted protein backbones, with the incorporation of unnatural amino acids (UAAs), alter protein topology, stability, and dynamics. First, we design knotted protein variants where specific residues within a knotted core are replaced with UAAs. Structural predictions are then performed using AlphaFold to assess the feasibility of the modified backbones. Lastly, molecular dynamics simulations are conducted in GROMACS to evaluate the folding behavior of the engineered proteins. By integrating topological analysis, AI-driven structure prediction, and molecular simulations we seek to understand fundamental principles dictating protein knotting. We anticipate that these findings will deepen our understanding behind knot stability and folding, laying the groundwork for future applications in de novo protein design.