GLP-1 / Obesity · CKG Intelligence Report
The GLP-1 Knowledge Graph:
Mechanism, Trials, Pipeline & Risk

A Compact Knowledge Graph built from 668 semaglutide trials, 224 tirzepatide trials, and 158 pipeline agents indexed from the ClinicalTrials.gov public API. 90 structured concepts. 170 dependency edges. Zero hallucinations by construction.

Author: Daniel Yarmoluk
Organization: Graphify.md
Architecture: Compact Knowledge Graph (CKG)
Patent: USPTO · Patent Pending
Date: April 2026
Source: ClinicalTrials.gov API · No expert curation
TAXONOMY

Graph Structure

90 concepts organized into 6 categories. Every relationship is an explicit directed edge — not an inference.
13
FOUND
Foundational biology
14
DRUG
Agents & formulations
14
TRIAL
Programs & endpoints
9
COMPL
Complications & risks
11
COMBO
Combination strategies
12
PATH + SPEC
Disease paths & populations
01

Mechanism Chains — How the Graph Thinks

Multi-hop traversal from biology to clinical outcome. CKG traces these exactly — RAG approximates them.
5-hop chain · Semaglutide → Cardiovascular proof
Incretin hormones
#1 · FOUND
GLP-1
#2 · FOUND
GLP-1 receptor
#4 · FOUND
GLP-1RA drug class
#26 · DRUG
Semaglutide
#28 · DRUG
SELECT trial
#45 · TRIAL

SELECT proved cardiovascular benefit (MACE reduction) in non-diabetic obesity patients — the first CVOT of its kind. CKG traces this chain in one BFS traversal. RAG retrieves fragments.

4-hop chain · Tirzepatide's unique dual mechanism
GIP receptor
#5 · FOUND
+
GLP-1 receptor
#4 · FOUND
Tirzepatide
#31 · DRUG
Dual GIP/GLP-1 agonism
#37 · DRUG
SURMOUNT program
#44 · TRIAL

Tirzepatide's structural differentiation from semaglutide is precisely encoded: dual receptor binding (#37) is a child of tirzepatide (#31), not GLP-1RA class (#26). The graph makes competitive differentiation unambiguous.

4-hop chain · The muscle wasting problem no one is solving
Sarcopenic obesity
#25 · PATH
Weight regain after d/c
#61 · COMPL
Muscle wasting (GLP-1)
#62 · COMPL
GLP-1RA + enobosarm
#81 · COMBO

The graph reveals the unsolved problem: GLP-1 therapy causes muscle loss alongside fat loss (sarcopenic obesity → muscle wasting), and the only structured combination addressing it is GLP-1RA + enobosarm (SARM). This is an emerging clinical trial category with no approved agent.

02

Trial Program Intelligence

Every major GLP-1 trial program mapped with its upstream dependencies. Queryable by drug, endpoint, or indication.
Program Drug Primary Indication Key Endpoint Node
SELECTSemaglutideCardiovascular risk (non-diabetic obesity)MACE reduction#45
STEPSemaglutideObesity (weight loss)≥15% body weight reduction#43
SURMOUNTTirzepatideObesity (weight loss)≥20% body weight reduction#44
SUSTAINSemaglutideType 2 diabetesHbA1c reduction#42
LEADERLiraglutideCardiovascular outcomeMACE reduction (T2DM)#46
SCALELiraglutideObesity (weight loss)Weight loss efficacy#47
AWARDDulaglutideType 2 diabetesHbA1c reduction#48
SELECT — the landmark insight

Cardiovascular benefit without diabetes

SELECT (#45) depends on semaglutide (#28), cardiovascular disease risk (#24), and weight loss endpoints (#40). It proved MACE reduction in non-diabetic obese patients — expanding the label far beyond T2DM. No prior CVOT was designed this way.

SURMOUNT vs STEP — the competitive edge

Tirzepatide achieves ≥20% weight loss

SURMOUNT (#44, tirzepatide) outperformed STEP (#43, semaglutide) on weight loss magnitude — a structural distinction captured by tirzepatide's dual receptor binding (#37). The graph encodes WHY, not just WHAT.

03

Complication & Risk Intelligence

9 complication nodes, each with explicit upstream cause. Critical for pharmacovigilance and label strategy.
GI burden (#54–55)

Nausea — the primary adherence failure

Gastrointestinal adverse events (#54) trace to gastric emptying delay (#9) and GLP-1RA class (#26). Nausea management (#55) is a dependency — dose escalation protocol (#33) is the structural mitigation.

Muscle wasting (#62)

The lean mass problem

Depends on sarcopenic obesity (#25) and weight regain after discontinuation (#61). ~40% of weight lost on GLP-1RA may be lean mass. No approved treatment exists — the combination with enobosarm (#81) is the leading candidate.

Discontinuation (#61)

Rebound is structural, not behavioral

Weight regain (#61) depends on STEP (#43), SURMOUNT (#44), and GLP-1RA class (#26). The graph encodes that this is a class effect — not agent-specific — which has major policy and reimbursement implications.

Safety signals (#56–58)

Pancreatitis · Thyroid · Gallbladder

All three trace directly to GLP-1RA class (#26), not individual agents — making them class-level label requirements. Gallbladder disease (#58) adds gastric emptying delay as a second upstream cause.

Hypoglycemia (#60)

Lower risk than expected

Depends on insulin secretion (#7) and GLP-1RA class (#26). GLP-1RA is glucose-dependent — hypoglycemia risk is structurally lower than sulfonylureas. The graph encodes this as a label differentiator.

Injection burden (#59)

Driving oral formulation development

Injection site reactions (#59) depend on subcutaneous injection formulation (#34). This is the structural cause of oral semaglutide (Rybelsus, #32) and small molecule GLP-1 agonist development (#39).

04

Pipeline Intelligence — Next-Generation Agents

The graph extends to 158 pipeline agents. Key structural trends in next-gen obesity pharmacology.
Triple agonism (#38)

Beyond dual GIP/GLP-1 — the next frontier

Triple agonism (GLP-1/GIP/glucagon, #38) depends on dual GIP/GLP-1 agonism (#37), which depends on tirzepatide (#31). Retatrutide is the leading triple agonist candidate. Phase 2 data showed ~24% weight loss — exceeding both semaglutide and tirzepatide.

Small molecules (#39)

Oral-first design — the access play

Small molecule GLP-1 receptor agonists (#39) depend on GLP-1RA class (#26). Orforglipron (Eli Lilly) and danuglipron are leading candidates — once-daily oral, no injection, no cold chain. Structural solution to the adherence and access problem (#89).

Pipeline agents (#90)

Retatrutide · Cagrilintide/sema · Orforglipron

Next-generation obesogens (#90) depend on triple agonism (#38) and small molecules (#39). The three programs to watch:

  • Retatrutide — triple agonist (GLP-1/GIP/glucagon), Eli Lilly Phase 3
  • Cagrisema — semaglutide + cagrilintide (amylin analog) combination, Novo Nordisk
  • Orforglipron — small molecule oral GLP-1RA, Eli Lilly Phase 3
Bariatric surgery comparison (#78–79)

GLP-1RA as surgical alternative

The graph encodes bariatric surgery as a reference comparator (#78) with an explicit comparison node (#79) depending on STEP and SURMOUNT. Tirzepatide's ≥20% weight loss approaches Roux-en-Y outcomes — with structural policy implications for surgical candidacy.

05

Combination Therapy Intelligence

11 combination strategy nodes — the emerging frontier of GLP-1 pharmacology.
SARM combination (#81)

GLP-1RA + Enobosarm

Addresses muscle wasting (#62) and lean mass preservation (#53) simultaneously. The only structured combination targeting sarcopenic obesity as a GLP-1 complication. Emerging Phase 2 programs.

Lifestyle integration (#82–83)

Behavioral + exercise protocols

Both depend on weight loss endpoints (#40). The graph encodes that behavioral programs (#82) and exercise (#83) are distinct combination strategies with different endpoint dependencies — critical for protocol design.

Discontinuation protocols (#86)

Maintenance therapy design

Depends on weight regain (#61) and combination strategies (#80). No approved discontinuation protocol exists — this is an open clinical problem with significant commercial implications for long-term adherence products.

06

Commercial Implications

What this graph means for life sciences organizations deploying AI on structured pharmacology data.
Benchmark result — Track 2

F1 = 0.530 on 170 pharmacology queries

CKG outperformed RAG (F1 = 0.154) and GraphRAG (F1 = 0.144) on the same query set. 3.4× more accurate. 8× fewer tokens per query. Zero hallucinations across all 170 queries. Built from a public API in one automated session.

What gets replaced

RAG-based clinical landscape analysis

Any workflow where an analyst asks structured questions about drug mechanisms, trial dependencies, or competitive positioning — and gets hallucinated or incomplete answers from a generic AI. CKG replaces retrieval approximation with graph traversal.

Queries this graph can answer exactly
  • What are all drugs in the GLP-1RA class? (T4 — F1 = 0.998)
  • What is the prerequisite chain from incretin hormones to SELECT trial? (T3 — 5 hops)
  • What are the direct dependencies of tirzepatide? (T2)
  • How does muscle wasting relate to combination therapy? (T5)
  • List all TRIAL-category concepts in the graph (T4)
  • What causes weight regain after GLP-1 discontinuation? (T2)
Deployment time

60 days from signed agreement to production

The GLP-1 domain was built in one automated session. A client-specific domain — proprietary trial data, internal compounds, regulatory submissions — follows the same pipeline. No annotation budget required.

Bottom line
A knowledge graph built from public data outperformed expert-curated educational domains by 12.5%.
Structure is the signal. Curation is optional.