The approach involves identifying patterns in chemical substructures and small molecules to guide drug discovery. By employing a module with the Monte Carlo tree search algorithm, similar to AlphaGo, the AI model can confidently identify the chemical substructures underlying predicted antibiotic activity. This explainable approach to AI allows for the identification of entire classes of compounds, rather than just single compounds. This method is different from traditional AI approaches in drug discovery and enables the discovery and description of new classes of antibiotics. The machine learning approach allows the generalization beyond the training data set and the identification of subtle patterns in chemical structures, which humans cannot do. This AI-powered approach is fundamentally different from the brute force method typically used by humans and enables the prediction of entire chemical substructures that define hundreds, if not thousands, of different chemicals.