Two questions: can we, and should we?
Every target faces two separate tests. The first is druggability: does this molecule have a pocket a drug can actually bind well? Enzymes and GPCRs are often druggable because evolution gave them neat pockets for their natural partners. The second is target validation: even if we can drug it, will hitting it actually help the patient? You need both a green light. A perfectly druggable target that doesn't matter to the disease is a trap, and a disease-critical target with no bindable pocket is a heartbreak.
- Druggable? Look for a well-formed pocket — a cleft the right size and chemistry for a small molecule to grip.
- Validated? Gather evidence — genetics, animal models, human biology — that changing this target changes the disease.
- Confirmed in the lab? Show target engagement — proof the drug actually binds the target inside living cells, not just in a test tube.
Hit it cleanly: selectivity
A great target also needs to be hittable cleanly. Selectivity is the drug's ability to bind its intended target while leaving close relatives untouched. This is hardest within big families — there are hundreds of kinases sharing similar active sites — so a kinase drug must thread a needle to avoid an off-target effect on the wrong relative. Poor selectivity narrows the therapeutic window, the gap between a helpful dose and a harmful one. Much of lead optimisation is really the patient pursuit of selectivity.
The hard cases — and new tricks
Some of biology's most important targets have long been called "undruggable" because they lack a tidy pocket. The classic example is a protein–protein interaction: two proteins meeting over a broad, flat surface with no cleft for a small molecule to grab. For decades these defeated medicinal chemists, and many remain very hard.
But the toolbox keeps growing. New modalities are reopening targets once thought hopeless — for instance a PROTAC doesn't need to inhibit a protein at all; it grabs the target with one hand and the cell's disposal machinery with the other, and the protein is destroyed. The lesson of this whole track is that a "good target" is partly fixed biology and partly a moving line drawn by our chemistry. Master the families, ask can we / should we / cleanly?, and you can read any drug program with a clear eye.