PGFS++：合成与多样性约束下的分子性质优化
PGFS++: Molecular Property Improvement under Synthesis and Diversity Constraints
浏览论文内容
中文总结 AI 辅助
PGFS++是一种感知合成的强化学习框架，通过优化反应模板与库存构建模块的应用，在改进分子目标性质的同时保留输出多样性，解决了PGFS+的奖励作弊问题。
中文摘要 AI 辅助
在早期药物发现中，优化药物相似性或结合亲和力等分子性质是一项反复开展的任务。然而，在无约束化学空间中优化得到的分子若无法合成，实际应用价值有限。Policy Gradient for Forward Synthesis（PGFS）是一种用于分子优化的感知合成强化学习方法，但它采用反应物嵌入预测导致反应物选择间接，我们的研究表明这会限制学习效果。我们首先开发了PGFS+，其中反应模板和第二反应物由可训练的嵌入查找表表示，结合更有效的评分函数和强化学习算法，PGFS+显著提升了目标性质。不过，它存在奖励作弊的失效模式：强大的反应物搜索可将不同输入分子映射到同一高奖励“磁体分子”，在提升奖励的同时导致输出多样性崩溃。因此，我们提出PGFS++，这是一种针对输入特定分子优化的感知合成强化学习框架。给定输入分子，PGFS++将其视为正向合成轨迹的起点，应用学习到的反应模板与兼容的库存构建模块，生成具有改进目标性质、明确合成路线且与输入结构相似的分子。分子优化任务实验表明，PGFS++在提升目标性质的同时保持了高输出多样性。
英文摘要
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.
发表机构
- Graphcore（格洛科普（Graphcore））
- University of Cambridge（剑桥大学）
机构由 AI 辅助整理，请以论文原文为准。
