The intelligence layer for uncompromising biopharma teams

DrugBank's Knowledge Graph is the connected intelligence core built from 20 years of curated biomedical data. Every datapoint traces back to its source, updates continuously, and is ready for AI and human discovery.

  • Drugs
  • Targets and enzymes
  • Conditions
  • Structured indications
  • Clinical trials
  • Companies
  • Adverse effects
  • Drug interactions
  • Pathways
  • Biomarkers
150M+ data points.
Twenty years of curated biomedical data, structured so a scientist and an agent get the same answer from the same question.
Billions of directional relationships.
The graph records a relationship's direction, not just that a connection exists. Inhibiting and inducing aren't the same.
Causal links, not co-occurrence.
Terms are connected when a mechanism is real, not because terms appear together in an abstract, a distinction most graphs skip.
Pipeline coverage of 10,000+ companies.
Every asset, phase, and sponsor is tied to the underlying biology, so commercial and scientific context show up in one query.
Traceable to primary source.
Every datapoint carries its evidence lineage, so the answer holds up when your VP, your partner, or a regulator asks where it came from.
Engineered to close the gaps.
Relationships are built in-house to fill the curation gaps left by the industry that stall hard and complex queries.

Explore with confidence. Decide with conviction.

One connected graph means completeness is guaranteed, not assumed.

  • Trust that whitespace is a finding, not missing data
  • Explore the complete scientific and commercial picture
  • Make go/no-go calls based on the full landscape
20-year
track record
1,500+
institutions
65,000+
citations
$10B+
in research enabled
10/20
top pharma companies

Knowledge Engine

Intelligence this nuanced isn't built overnight

Run continuously and overseen by expert curators, the Knowledge Engine delivers complete, correct, fresh, and traceable data directly to the graph.

Capture & Ingest

Continuous data point ingestion from a broad set of primary and regulatory sources

Engineer & Structure

Aggregate, harmonize, and resolve entities to a single record and connect to 20+ ontologies

Validate & Review

Ongoing quality control overseen by biostatisticians to maintain standards for millions of records

Engineered beyond the industry's blind spots

Sponsors resolved to their true owner

A sponsor's name on a filing isn't always who's behind it, and untangling that takes more than one lookup.

  • One company can rename itself multiple times over a trial's history.
  • One name can front a dozen affiliates, each filing under a different label.

Trial registries only ever show a piece of this. DrugBank pulls the rest from SEC filings, corporate registries, and M&A records, then normalizes and connects them, drawing on SEC records that reach back over a century.

Get answers you can trust, every time.

Claude · drugbank-mcp
Which approved GLP-1R agonists have a half-life long enough for weekly dosing?
drugbank-mcp · structured pharmacology · 11 approved, 9 with a published half-life
Three. The class runs from 26 minutes to 168 hours, and the labels report it in minutes, hours and days, so I sorted on the normalized value.
Half-life · approved GLP-1R agonists 9 of 11
Drug As published Normalized
Glucagon 26 minutes 0.43 h
Pramlintide 48 minutes 0.80 h
Exenatide 2.4 hours 2.4 h
Liraglutide 13 hours 13 h
Weekly dosing threshold · 24 h
Dulaglutide 3.75 days 89.9 h
Tirzepatide 5 days 120 h
Semaglutide 168 hours 168 h
Avexitide is an antagonist at this receptor. Excluded. · Release 1423972