Skip to content
Learn Motion
ExploreHow it worksMembership
Log in
Learn Motion

Can AI Discover New Drugs? A High-Level Overview

1Why Drug Discovery Is Hard, and Where AI Fits2How AI Learns From Molecules and Proteins3Finding and Validating a Biological Target4Designing Molecules: Generative AI and Virtual Screening5From Hit to Lead: Optimizing Properties With AI6What AI Still Cannot Do7Judging the Claims: Real Successes, Failures, and Open Questions
Finding and Validating a Biological Target

Biologically Valid but Practically Undruggable

3 / 4
Compare the two surfaces. On the left, the protein has a defined pocket — a cavity where a drug-sized molecule can sit, make several contacts, and be held in one orientation. Enzymes and receptors usually look like this, and most approved small-molecule drugs act on targets of this kind. On the right, the relevant surface is broad and flat. There is nowhere for a small molecule to grip, so getting both strong binding and selectivity becomes very hard. The key point is that this is a structural question, not a biological one: the protein on the right can be central to the disease and still be a poor drug target. And tractability is a current assessment, not a permanent label — targets once called undruggable have become tractable when a previously hidden pocket was revealed.
0:00 / 0:00

A target can be genuinely involved in a disease and still be a poor drug target, because druggability — also called tractability — is a separate question from biological validity. Druggability asks whether a small molecule can bind the protein with enough affinity and specificity to change its behavior, and the answer depends heavily on the protein's three-dimensional surface.

Proteins with a well-defined pocket — a cavity or groove where a small molecule can sit, make several contacts, and be held in a specific orientation — are generally tractable. Enzymes with a catalytic site and receptors with a ligand-binding site are the classic examples, and most approved small-molecule drugs act on targets of this kind. Proteins whose relevant surface is broad, flat, and featureless offer few anchor points; a small molecule has little to grip, and achieving both strong binding and selectivity against similar proteins becomes very difficult.

The distinction is structural, not biological. The protein may be central to the disease and still present no usable pocket. This is why the sequence-versus-structure distinction matters here: knowing a target's sequence tells you what the protein is, but only the folded structure — or a reliable predicted structure — tells you whether a drug-sized molecule has somewhere to bind. Some historically "undruggable" targets became tractable once new modalities or newly resolved structures revealed a pocket that had not been visible before, which is a reminder that tractability is a current assessment, not a permanent property.

Previous3 / 4Next

Learn Motion

Generate a course. Learn it properly.

Operated by Wuhan Daoyin Technology Co., Ltd.

Contact: [email protected]
Privacy PolicyTerms of Service

© 2026 Learn Motion