Fig. 3: Generated molecules over five outer AL cycles for CDK2.
From: Optimizing drug design by merging generative AI with a physics-based active learning framework
A Cumulative plot of the specific set enrichment over the inner/outer AL cycles. Vertical dotted lines mark the end of each outer AL cycle, dividing the plot into separate outer cycles. Red dots represent inner AL cycles using the CDK2 initial-specific set. Blue dotted lines indicate the integration of permanent-specific sets from the third and fourth outer AL cycles by the fifth outer AL cycle. B UMAP illustrating the generated molecules over the outer AL cycles, with marker styles and colors indicating different cycles as shown in the legend. The initial-specific set is labeled ‘initial’ and outer AL cycles as “outerX” C Glide gscore histograms of docked molecules across outer AL cycles compared to the Glide gscore histogram of the initial-specific set of CDK2. A red vertical line marks the −8.0 kcal·mol−1 gscore threshold. D Zoomed-in view of the best scoring molecules in (C).
