>>11159064So most of the time they are based on crystal structures of the actual ligand, then they try to make that into the smallest or most inert molecule possible while still retaining the high binding affinity. or adjust that affinity, ect.
But nothing novel novel has every been designed to my knowledge. You have to work backwards here, without knowing at the very least what a receptor binding pocket looks like.
But then also algorithms which optimize the thermodynamics of candidate molecules with the receptor pocket help, to screen out tons upon tons of potential molecules. Basically molecular dynamics but taking into account individual atom affinities and shit, I'm talking way beyond my scope here though
