The paper explores how large language models (LLMs) can be trained for small-molecule drug design by using synthetic tasks that are cheaper to evaluate. By employing a curriculum that gradually increases task difficulty, the authors demonstrate that LLMs can learn design strategies that outperform larger models on structure-based lead optimization. This approach shows that scaling post‑training with synthetic tasks can effectively adapt LLMs to high‑cost experimental scenarios that are otherwise infeasible to train on directly.
arXiv:2607. 19044v1 Announce Type: new Abstract: Leveraging large language models (LLMs) for molecular generation has shown remarkable potential in chemical and drug design.
arXiv:2606. 11256v1 Announce Type: cross Abstract: Designing molecules with target properties is most useful when candidate structures are accompanied by feasible synthetic routes.
arXiv:2603.03517v2 Announce Type: replace-cross Abstract: General-purpose large language models (LLMs) that rely on in-context learning do not reliably deliver the scientific understanding and perfor...
The paper introduces Round-Trip Reinforcement Learning (RTRL), a framework that trains chemical language models to improve round‑trip consistency by rewarding successful forward and reverse transformations. By iteratively training forward and reverse mappings, RTRL leverages abundant unlabeled chemical data to enhance both consistency and overall performance across supervised, self‑supervised, and synthetic data regimes. Experiments show that RTRL outperforms strong baselines, demonstrating that round‑trip consistency can be treated as a trainable objective for more robust foundation models.
arXiv:2607. 02834v1 Announce Type: new Abstract: Molecular optimization often starts from a pretrained generative model that captures a broad prior over valid molecular structures.
arXiv:2603. 12666v2 Announce Type: replace-cross Abstract: Retrosynthesis prediction aims to identify reactants that can synthesize a given product molecule.
arXiv:2604.07669v3 Announce Type: replace-cross Abstract: Synthesizable molecular optimization seeks to improve target properties while ensuring that molecular modifications follow feasible synthetic...
arXiv:2605. 15354v2 Announce Type: replace Abstract: Despite the success of foundation models in language and vision, molecular graph generation still lacks a unified framework for heterogeneous design tasks with reliable controllability.
arXiv:2606. 03057v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used for molecular tasks, but it remains unclear which molecular representation to use.
PGFS++ is a synthesis‑aware reinforcement learning framework that improves molecular properties such as drug‑likeness or binding affinity while ensuring the resulting molecules can be synthesized and remain structurally similar to the input. It builds on PGFS+ by using trainable embedding lookup tables for reaction templates and second reactants, a more effective scoring function, and a refined RL algorithm. The method addresses a reward‑hacking failure mode by treating each input molecule as the start of a forward‑synthesis trajectory, applying learned reaction templates with in‑stock building blocks, and producing diverse, high‑quality outputs with explicit synthesis routes.
PGFS++ is a synthesis‑aware reinforcement learning framework that improves molecular properties while ensuring the resulting molecules can be synthesized and remain structurally similar to the input. It builds on PGFS+ by using trainable embedding lookup tables for reaction templates and second reactants, a more effective scoring function, and a refined RL algorithm. Experiments demonstrate that PGFS++ enhances target properties and preserves high output diversity, overcoming the reward‑hacking failure mode seen in earlier versions.
The study evaluates four pretrained molecular language models on six virtual libraries covering drug discovery, organic materials, and catalysis. It finds that native embeddings vary widely in performance, while molecular fingerprints remain consistently strong. Fine‑tuning the models on library‑specific data markedly improves sample efficiency, with several adapted encoders outperforming others across all tasks.