← All Targets

Amyloid beta

4 drugs

CNS

Approved Drugs

Companies

Indications

Therapeutic Areas

Broadest Approval

KISUNLA

Eli Lilly

approved indications

Human Genetic Evidence Strong

Genetic Verdict

✅ STRONG SUPPORT

Clinical Translation ⓘ

~1.8x

vs baseline success

Direction

🎯 Inhibition likely beneficial

Confidence

Low (50% consistent)

Key Risks

⚠ Mixed direction signals

💡 Why inhibition?

• Loss-of-function variants reduce disease risk (OR < 1)
• 50% directional consistency across 2 traits
• Strong signal in psychiatric disorder, nervous system disease, phenotype pathways

Cross-Disease Effects

Trade-off: High

⚠ psychiatric disorder → inhibition may be harmful

⚠ nervous system disease → inhibition may be harmful

✔ phenotype → inhibition beneficial

⚠ genetic, familial or congenital disease → inhibition may be harmful

⚠ nutritional or metabolic disease → inhibition may be harmful

⚠ cardiovascular disease → inhibition may be harmful

⚠ 5 area(s) show opposite direction — consider disease-specific targeting

⚠ High trade-off: inhibition helps some diseases but may worsen others. Consider disease-specific approaches.

Direction of Effect

50% aligned

✔ Alzheimer disease → Inhibition

✗ cerebral amyloid angiopathy, APP-related → Activation

Evidence Across Diseases

20 total

Strong

Moderate

Weak

GWAS and other genetic studies link APP to 34 diseases.

Alzheimer disease type 1

nervous system disease, genetic, familial or congenital disease

0.95

association score

Alzheimer disease

psychiatric disorder, nervous system disease

0.87

association score

Loss-of-function protective

Inhibiting this target may be therapeutic

cerebral amyloid angiopathy, APP-related

psychiatric disorder, nervous system disease

0.86

association score

Loss-of-function risk

Loss-of-function causes disease; activation may help

Understanding these scores

Association Score (0-1): Combines all evidence types (genetic, literature, drugs, animal models). Higher = more evidence linking target to disease. This is a ranking heuristic, not a confidence score.

L2G Score: Open Targets uses a machine learning model (Locus-to-Gene) to predict which gene is causal at each GWAS locus. L2G=0.5 means ~50% probability of being the causal gene. Only associations with L2G > 0.05 are included.

Odds Ratio (OR): From gene burden studies (UK Biobank, AstraZeneca PheWAS). Measures how loss-of-function variants affect disease risk. OR<1 = protective (inhibiting target may help), OR>1 = risk (losing function causes disease).

Beta (β): Effect size for continuous traits. β<0 = protective, β>0 = risk.

Clinical Translation (~1.8x): Based on Nelson et al. 2015: drug targets with genetic evidence have ~2x higher success rates in clinical trials. We estimate: Strong support (score ≥0.7) → ~1.8x, Moderate (0.3-0.7) → ~1.3x, Weak → baseline.

Colocalization (H4): Tests whether a GWAS signal and an eQTL/pQTL signal share the same causal variant. H4 is the posterior probability that both traits are associated AND share a causal variant. H4 > 0.8 = strong evidence that gene expression/protein levels drive disease risk. This links genetic variation → gene expression → disease, supporting the target-disease connection.

Source: Open Targets Platform · Updated 2026-04-01