Domain-Adapted Molecular Language Models for Efficient Search of Make-on-Demand Libraries

This dataset contains all relevant data for the manuscript "Domain-Adapted Molecular Language Models for Efficient Search of Make-on-Demand Libraries", including `code`: A snapshot of the `bayleys` Github repository (V.1.0.0). `libraries`: All labeled molecule libraries used in the performed benchmarking studies, as well as corresponding metadata and data provenance. Libraries can be directly read using the `bayleys` Python package. `configs`: All experiment configurations used for performing the experiments described in the paper, including encoder configurations, surrogate model configurations, acquisition configurations, and pre-training configurations. Configurations are provided in the format required by the `bayleys` library. `results`: Optimization trajectories of all experiments described in the paper. Trajectories were directly generated using the `bayleys` library.