Search papers, labs, and topics across Lattice.
This paper introduces PGFS++, an advanced synthesis-aware reinforcement learning framework designed to enhance molecular properties while adhering to synthesis and diversity constraints. By utilizing learned reaction templates and in-stock building blocks, PGFS++ effectively generates improved molecules with explicit synthesis routes, addressing limitations found in its predecessor, PGFS. Experimental results demonstrate that PGFS++ not only enhances target properties but also maintains high output diversity, crucial for practical applications in drug discovery.
PGFS++ achieves significant molecular property improvements while ensuring diverse outputs, overcoming the reward-hacking pitfalls of previous methods.
Improving molecular properties, such as drug-likeness or binding affinity, is a recurring task in early-stage drug discovery. However, molecules optimized in an unconstrained chemical space have limited practical value if they cannot be synthesized. Policy Gradient for Forward Synthesis (PGFS) is a synthesis-aware reinforcement learning method for molecular improvement, but its use of reactant embedding prediction makes reactant selection indirect, which, as we show, limits learning effectiveness. We first develop PGFS+, in which reaction templates and second reactants are represented by trainable embedding lookup tables. Combined with a more effective scoring function and RL algorithm, PGFS+ significantly improves the desired property. However, it exposes a reward-hacking failure mode: a powerful reactant search can map diverse input molecules to the same high-reward magnet molecule, improving the reward while collapsing the output diversity. We therefore introduce PGFS++, a synthesis-aware reinforcement learning framework for input-specific molecular improvement. Given an input molecule, PGFS++ treats it as the start of a forward-synthesis trajectory, applies learned reaction templates with compatible in-stock building blocks, and produces a molecule with improved target properties, an explicit synthesis route, and structural similarity to the input. Experiments on molecular improvement tasks show that PGFS++ improves target properties while preserving high output diversity.
