Finding Drug Candidate Hits With A Hundred Samples: Ultralow Data Screening With Active Learning

Active learning (AL) can significantly accelerate drug discovery by iteratively selecting informative molecules, reducing experimental workload. However, existing AL studies typically assume access to large datasets, an unrealistic scenario for most academic labs. AL strategies tailored specifically for small‐scale molecular screening, are investigated, using only 110 affinity evaluations approximated by docking scores from realistic compound libraries: the Developmental Therapeutics Program repository (DTP) and Enamine Discovery Diversity Set 10 (DDS‐10). Among 20 tested combinations of molecular descriptors and machine learning models, continuous and data‐driven descriptors combined with a multilayer perceptron, augmented by the pairwise difference regression data augmentation technique, are identified as optimal. This combination achieves a 100% probability of discovering at least five top‐1% hits from DTP using only 110 affinity evaluations which remains high under simulated experimental uncertainty. Similarly, the DDS‐10 dataset achieves a 100% probability of discovering at least five top‐1% hits. Incorporating prior knowledge by enriching initial datasets with a single known hit molecule increased the probability of finding 20 or more hits. These findings underscore the feasibility and substantial potential of AL for small‐scale drug discovery in resource‐limited environments. These results suggest that early in the AL search the algorithm benefits from accurately quantifying the binding strengths of very weak binders.