A.1 Additional experimental results We further introduce our additional experiments in this section. Standard deviation is given in brackets. Significance analysis of the results presented in Figure 5 in the main text. We compared the models' performances in terms of hit The result is shown in Table 4. From the Notably, the comparison with FREED(PE) results in a p-value of 0.0066, even below the Fragments from known active compounds. In Figure 1, active fragments-generated molecules show higher docking scores than random fragments-generated molecules.

Deep reinforcement learning has successfully been applied for molecular discovery as shown by the Molecule Deep Q-network (MolDQN) algorithm. This algorithm has challenges when applied to optimizing new molecules: training such a model is limited in terms of scalability to larger datasets and the trained model cannot be generalized to different molecules in the same dataset. In this paper, a distributed reinforcement learning algorithm for antioxidants, called DA-MolDQN is proposed to address these problems. State-of-the-art bond dissociation energy (BDE) and ionization potential (IP) predictors are integrated into DA-MolDQN, which are critical chemical properties while optimizing antioxidants. Training time is reduced by algorithmic improvements for molecular modifications. The algorithm is distributed, scalable for up to 512 molecules, and generalizes the model to a diverse set of molecules. The proposed models are trained with a proprietary antioxidant dataset. The results have been reproduced with both proprietary and public datasets. The proposed molecules have been validated with DFT simulations and a subset of them confirmed in public "unseen" datasets. In summary, DA-MolDQN is up to 100x faster than previous algorithms and can discover new optimized molecules from proprietary and public antioxidants.