From a structure to a design map
Structure-based drug design starts with a 3D picture of the target — usually a protein — solved by X-ray crystallography or cryo-EM, ideally with a ligand already bound. Your job is to read that picture the way a climber reads a rock face: where are the holds, which ones are solid, and where can a new molecule grip? The single most important feature is the binding pocket (often the enzyme's active site): a cavity in the protein surface where small molecules can sit.
A pocket only matters if a drug-like molecule can actually bind it tightly enough — that property is called druggability. Deep, hydrophobic, well-defined pockets tend to be druggable; flat, exposed, or highly polar surfaces (think many protein–protein interactions) are far harder. Reading a structure is partly about judging whether the pocket in front of you is a fair fight.
The interactions that hold a ligand in place
Molecular recognition is the sum of many weak, specific contacts. The vocabulary you must recognize in any structure: a hydrogen bond between a donor and acceptor a few ångströms apart; the hydrophobic effect that buries greasy surfaces away from water; van der Waals packing where shapes match; a salt bridge or ionic interaction between charges; and aromatic π-stacking. Each one is a place where a substituent could be added, kept, or improved.
Not all contacts are equal. A binding hotspot is a small sub-region of the pocket that contributes a disproportionate share of the affinity — often a deep hydrophobic spot or a buried polar anchor. If your molecule reaches a true hotspot, you win potency cheaply; if it only paints the easy, solvent-exposed rim, you will struggle. Mapping hotspots first tells you where to spend your synthetic effort.
Turning the map into design ideas
- Identify the pocket and confirm it is real and druggable — check the bound ligand, the density, and whether the cavity is enclosed.
- List the existing ligand's key interactions one by one, naming each (H-bond, salt bridge, π-stack) and the residue involved.
- Find the hotspots and any empty sub-pockets — places where the current molecule does not yet reach.
- Propose specific edits: a group to fill an empty pocket, an H-bond donor to pair with an acceptor, a ring to stack — then test them computationally before synthesizing.
This is the loop the rest of the track teaches you to run faster and more reliably: read the structure, propose an edit, score it, and decide what to make. The structure does not design the molecule for you — it tells you which of your ideas are even plausible, and it kills bad ideas cheaply.