AI summaryⓘ
The authors address the challenge of designing molecules that meet certain properties by improving how the search process is guided. Instead of relying on standard methods that use simplified molecule representations, they use a language model to understand the design task and feedback in a natural text form. Their method, called method, helps pick promising reference molecules to explore or exploit the search space more effectively, updating instructions for the molecule generator each round. Tests show that their approach performs better than simple one-step prompting and is competitive with traditional optimization methods. They also find the best strategy varies by task type, like drug discovery or materials design.
Molecular inverse designBayesian optimizationGaussian processSMILESLarge language modelClosed-loop optimizationOracle feedbackMolecule generationExploration-exploitationDrug discovery
Authors
Yaoyao Xu, Xinjian Zhao, Xiaozhuang Song, Lei Bai, Tianshu Yu
Abstract
Practical molecular inverse design is rarely a one-shot generation problem; it often takes the form of closed-loop candidate-pool enrichment, where under a limited oracle budget the goal is to \emph{increase the fraction of generated molecules that match a desired property profile}. Bayesian optimization (BO) offers a natural framework for this setting, yet standard Gaussian-process surrogates typically operate in compressed continuous embeddings, which discard the substructural and reference-similarity signals that chemists naturally use to decide where to look next. We propose \textbf{\method}, a closed-loop framework in which the surrogate, rather than the generator, is treated as the locus of design choice, and instantiate it with a frozen large language model that reasons directly over the task instruction, SMILES-level optimization history, and oracle feedback in their native textual form. At each iteration, the surrogate returns a structured decision signal that selects informative reference molecules under an exploration and exploitation principle, optionally with a concise guidance sentence. This signal is converted into next-round conditioning text for a frozen molecular generator, yielding an inspectable optimization trace in natural language. Experiments on MolQA drug and material design tasks show that \method improves over one-shot prompting, is competitive with or stronger than GP-based BO baselines, and reveals a domain-dependent interface: reference-only transfer works best for binary drug targets, while adding a concise surrogate summary is more beneficial for continuous material