Molecular function-guided unsupervised similarity-based enrichment for low-data virtual screening

Identifying bioactive compounds for orphan or understudied protein targets remains a significant hurdle in drug discovery due to the lack of sufficient structure-activity relationship (SAR) data. We present MolFuSE (Molecular Function-guided Similarity Explorer), an unsupervised, projection-based enrichment framework designed for low-data virtual screening and ligand repurposing. MolFuSE leverages manifold learning-specifically uniform manifold approximation and projection (UMAP)-to create low-dimensional similarity spaces based on a reference set of ligands known to bind proteins with the same molecular function as the target. Through systematic evaluation across nine protein targets and their functional classes, we demonstrate that UMAP embeddings of Mordred 2D descriptors act as a strong "potency filter", significantly outperforming traditional topological fingerprints in the early enrichment of high-potency inhibitors (<100 nM). While extended-connectivity fingerprints (ECFP4) fingerprints achieve superior global ranking (receiver operating characteristic (ROC)-area under the curve (AUC) 0.961 vs. 0.863), the descriptor-based approach provides more robust early recognition (enrichment factor at 1% (EF@1%)) when the reference set itself is small. Crucially, we identify a critical point in data scaling: performance peaks at a reference set size of approximately 10,000 compounds and degrades for larger sets due to descriptor-space dilution from structurally diverse weak binders. MolFuSE provides a computationally efficient, interpretable tool for prioritizing high-affinity candidates when target-specific data is unavailable.