The Historical Accident That Split Drug Design in Two (And the Contrastive Model That Reunites It)
•12 min read
#drug discovery#contrastive learning#computational biology#virtual screening#protein-ligand interactions#graph neural networks#machine learning
Structure-based and ligand-based drug design evolved as separate fields solving the same problem. ConGLUDe, a contrastive geometric learning model, unifies both approaches and outperforms specialist methods on realistic benchmarks without requiring pre-defined binding pockets.