GFlowNet Pretraining with Inexpensive Rewards
This work addresses the challenge of generating diverse drug-like molecules for pharmaceutical research, representing an incremental advancement by combining existing GFlowNet concepts with novel pretraining and fine-tuning strategies.
The paper tackled the problem of limited chemical space exploration in molecular generation by introducing Atomic GFlowNets (A-GFNs), which use individual atoms as building blocks and are pretrained with inexpensive molecular descriptors, resulting in improved performance over baselines in drug design tasks.
Generative Flow Networks (GFlowNets), a class of generative models have recently emerged as a suitable framework for generating diverse and high-quality molecular structures by learning from unnormalized reward distributions. Previous works in this direction often restrict exploration by using predefined molecular fragments as building blocks, limiting the chemical space that can be accessed. In this work, we introduce Atomic GFlowNets (A-GFNs), a foundational generative model leveraging individual atoms as building blocks to explore drug-like chemical space more comprehensively. We propose an unsupervised pre-training approach using offline drug-like molecule datasets, which conditions A-GFNs on inexpensive yet informative molecular descriptors such as drug-likeliness, topological polar surface area, and synthetic accessibility scores. These properties serve as proxy rewards, guiding A-GFNs towards regions of chemical space that exhibit desirable pharmacological properties. We further our method by implementing a goal-conditioned fine-tuning process, which adapts A-GFNs to optimize for specific target properties. In this work, we pretrain A-GFN on the ZINC15 offline dataset and employ robust evaluation metrics to show the effectiveness of our approach when compared to other relevant baseline methods in drug design.