Optimisation of Molecular Binding Affinity in Latent Space with an Hybrid Generative-Evolutionary Framework

Key Points

• The research aims to optimize molecular binding affinity leveraging a hybrid generative-evolutionary framework.
• Merged results from benchmarking experiments were collected for the TED Commons DRD3 docking objective.
• Generated molecules were evaluated using parameters such as SMILES string, docking score, and generation number.
• Data supports reproducible analysis of docking-score trajectories across different generations and configurations.
• Generated docking scores allow for tracking molecular binding changes across iterations.
• Visualization of docking-score trajectories demonstrates the effectiveness of various model configurations.