基于AlphaFold 3与分子动力学模拟解析蛋白-核酸适配体的识别机制

ptamers, as emerging molecular recognition elements, demonstrate tremendous potential in targeted therapy and molecular diagnostics due to their high stability, ease of synthesis, programmability, and chemical modifiability. However, their functionality critically depends on specific three-dimensional conformations. While conventional experimental approaches face limitations in resolving nucleic acid-protein complex structures, existing computational methods suffer from insufficient accuracy and limited capability in modeling complex nucleic acid architectures. In this study, we employed the c-Met protein and its specific aptamer as a model system to innovatively integrate AlphaFold 3 with molecular dynamics (MD) simulations, enabling an in-depth investigation of their molecular recognition mechanism.First, the three-dimensional structure of the aptamer was predicted using AlphaFold 3 and subsequently validated through MD simulations. Next, molecular docking coupled with MD simulations was performed to identify the dominant binding mode, followed by residue energy decomposition to pinpoint critical nucleotide interaction sites. Finally, these key sites were experimentally verified using biolayer interferometry (BLI) and circular dichroism (CD) spectroscopy.To our knowledge, this study represents the first comprehensive framework that seamlessly integrates static structure prediction, dynamic interaction analysis, and experimental validation. Our findings provide crucial theoretical insights for developing next-generation aptamer-based targeted therapeutics and advanced molecular diagnostic technologies.