phenotypic screening
Suppose you want to find something that cures a sick cell, but you do not yet know which molecule inside the cell is the culprit. Phenotypic screening asks the practical question directly: 'Does this compound make the cell get better?' — judging compounds by the visible outcome (the phenotype) rather than by hitting a chosen target. It is testing the medicine by whether the patient improves, not by which knob it turns.
In a phenotypic screen, compounds are applied to living cells, tissues, or whole organisms, and you measure a biologically meaningful change: cells dying, a parasite clearing, a marker switching on, or a shape changing under a microscope. Because the readout reflects the full living system, a hit is already known to do something useful in context, even though you do not yet know how.
The strength of phenotypic screening is that it can find drugs working through unexpected mechanisms or several targets at once — historically a productive source of first-in-class medicines. The catch is the flip side: a hit's mechanism of action is unknown, and figuring out which target is responsible (target deconvolution) can be slow and hard, which complicates optimization and safety assessment.
Compounds are added to malaria-infected red blood cells and scored for killing the parasite; potent hits are pursued even though the molecular targets are initially unknown.
A phenotypic screen rewards a real biological outcome before any target is known.
Phenotypic screening sells biological relevance for mechanistic clarity: you know the compound works in cells, but not yet why — and the optimization can stall until the target is identified.