研究業績

This study presents a novel in silico screening method for identifying potential protein-ligand interactions within the complete structural proteome, as predicted by AlphaFold2, using puberulic acid as a test case. The methodology utilizes molecular docking simulations to predict interactions based on structural compatibility and binding affinity estimations. This approach highlights proteins, such as sodium/myo-inositol cotransporters, which exhibit theoretical high affinity for puberulic acid, suggesting possible competitive inhibition or mimetic interactions that could disrupt osmoregulation. It is important to clarify that these simulations provide a hypothetical framework for understanding potential interactions. The high docking scores do not directly translate to biological effects but serve as a basis for further experimental validation. This cautious interpretation acknowledges the inherent limitations of predictive modeling while maintaining confidence in the robustness of the computational process. Moreover, the exploratory nature of this study is supported by the fact that binding affinity predictions are not influenced by protein amino acid length or the predictive accuracy (pLDDT scores) provided by AlphaFold2. This independence from structural prediction metrics suggests that the docking results offer a reliable exploration of potential interactions across a wide array of proteins, making this approach a valuable asset in the fields of drug discovery and toxicological research. The ability to comprehensively identify candidate proteins with potential affinity to chemical compounds enhances the predictive capacity of toxicological assessments and supports the strategic development of safer pharmaceutical agents.