跳到主要内容
Learn Motion
探索课程产品能力价格

下载 Learn Motion

用手机扫描二维码,下载 App。

App Store

扫码从 App Store 下载 Learn Motion打开商店页面

Google Play

扫码从 Google Play 下载 Learn Motion打开商店页面
登录
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

Turning Scattered Evidence Into a Ranked List

2 / 4
Look at the three streams on the left. Human genetics carries the most causal weight: if a variant consistently protects people from a disease, that points to the gene being involved, not just correlated. Literature is broad but uneven — heavily studied proteins look important partly because they were studied. Omics shows where a gene is active and how it shifts across tissues, but that is association, not cause. The model's job is to merge these into one ranked list, weighting each stream by how much causal evidence it tends to carry. Watch what the ranking does not do: it does not confirm biology. A target can rank low simply because it has been under-studied, and a high rank means the evidence converged, not that the hypothesis is proven.
0:00 / 0:00

No single source of evidence settles whether a protein causes a disease. Human genetics offers the strongest signal: if people who carry a particular variant of a gene are consistently protected from or predisposed to a disease, that is evidence the gene's protein is causally involved, not merely correlated with it. Literature records decades of experimental observations, but it is uneven — heavily studied proteins accumulate evidence simply because they were studied. Omics measurements (transcriptomics, proteomics, and similar large-scale readouts) show where a gene or protein is active and how it changes across tissues and disease states, but they describe association, not causation.

AI's contribution at this stage is integration. A model takes these heterogeneous inputs — genetic association statistics, literature-derived relationships, expression and abundance measurements, pathway membership — and produces a single ranked list of candidate targets, weighting each source by how much causal evidence it tends to carry. The ranking is a prioritization, not a measurement: it says which hypotheses are worth testing first, given everything currently known.

The output inherits the weaknesses of its inputs. Literature-derived features are biased toward well-studied proteins, so a genuinely important but under-studied target can rank low. Genetic evidence is strongest where large biobanks and genome-wide association studies exist, which skews toward populations that have been sampled. A high rank means the evidence converged; it does not mean the biology is confirmed.

上一页2 / 4下一页

Learn Motion

生成课程,认真学会。

运营方:Wuhan Daoyin Technology Co., Ltd.

联系我们: [email protected]
隐私政策用户服务协议

© 2026 Learn Motion