Why potency alone fools you
It is almost always possible to make a molecule bind harder by making it bigger and greasier — bolt on another ring, add a lipophilic tail, fill more of the pocket. The IC50 improves and everyone celebrates. But if every potency gain comes from added size and lipophilicity, you are quietly building a molecule that will be insoluble, metabolically unstable, and promiscuous. This slow drift is sometimes called molecular obesity.
Efficiency metrics fix this by asking not "how potent?" but "how potent per unit of cost?" The two costs that matter most are molecular size and lipophilicity, which gives us two metrics.
Ligand efficiency: potency per atom
Ligand efficiency (LE) normalizes binding energy by the number of heavy (non-hydrogen) atoms. Roughly, LE = 1.37 × pIC50 / (heavy-atom count), in kcal/mol per atom. A small fragment that binds modestly can have excellent LE; a huge molecule that binds tightly can have poor LE. LE tells you whether each atom you add is earning its keep.
Compound A: IC50 = 100 nM (pIC50 = 7.0) heavy atoms = 18 LE_A = 1.37 x 7.0 / 18 = 0.53 kcal/mol per atom (good) Compound B: IC50 = 10 nM (pIC50 = 8.0) heavy atoms = 38 LE_B = 1.37 x 8.0 / 38 = 0.29 kcal/mol per atom (poor) B is 10x more potent but far less efficient: the extra 20 atoms bought only one log of potency.
Lipophilic efficiency: potency net of grease
Lipophilic ligand efficiency (LipE, also written LLE) is even simpler and arguably the single most useful number in optimization: LipE = pIC50 − logP (or logD). It rewards potency that comes from specific interactions — a real hydrogen bond, a salt bridge, good shape fit — and penalizes potency that comes merely from being greasy. Most good oral drugs sit around LipE ≥ 5, and a healthy series shows LipE rising over time.