The simulation visualizes a candidate molecule's eight molecular properties as individual desirability curves in 3D, showing how each raw property value maps to a 0-to-1 desirability score and how those scores combine via a geometric mean into the final QED value.
Adjust sliders for properties like molecular weight, LogP, hydrogen bond donors and acceptors, and rotatable bonds to reshape the molecule, and watch each desirability curve update in real time along with the combined QED score at the top.
Molecular property sliders (weight, LogP, HBD, HBA, PSA, rotatable bonds, aromatic rings), structural alert toggle, reset molecule
In the original 2012 study, Bickerton and colleagues fitted QED's desirability functions using property distributions from 771 oral drugs approved by regulatory agencies, so the curves reflect empirical patterns in real successful medicines rather than arbitrary cutoffs.
The simulation visualizes a candidate molecule's eight molecular properties as individual desirability curves in 3D, showing how each raw property value maps to a 0-to-1 desirability score and how those scores combine via a geometric mean into the final QED value.
The simulation visualizes a candidate molecule's eight molecular properties as individual desirability curves in 3D, showing how each raw property value maps to a 0-to-1 desirability score and how those scores combine via a geometric mean into the final QED value.
Adjust sliders for properties like molecular weight, LogP, hydrogen bond donors and acceptors, and rotatable bonds to reshape the molecule, and watch each desirability curve update in real time along with the combined QED score at the top.
In the original 2012 study, Bickerton and colleagues fitted QED's desirability functions using property distributions from 771 oral drugs approved by regulatory agencies, so the curves reflect empirical patterns in real successful medicines rather than arbitrary cutoffs.