Healthcare & Life Sciences

Zero-ETL graph analytics for healthcare and life sciences.

Explore drug and target relationships, identify research cohorts, and trace treatment and prescribing networks over your existing data without loading a separate graph database.

acts onassociated withDrug ADrug BTarget 1Target 2DiseasePuppyGraphExisting clinical & research dataDrugs · Targets · Patients · Providers · Claimsacts onassociated withDrug ADrug BTarget 1Target 2DiseasePuppyGraphExisting clinical & research dataDrugs · Targets · PatientsProviders · Claims

01 / Graph analytics for healthcare & life sciences

Drug discovery, patient cohorts, and prescribing networks.

Follow connected records to examine drug mechanisms, find patients who match research criteria, and review treatment pathways and prescribing networks.

Examples use simplified graph models. Adapt labels and properties to your schema.

01

Drug discovery & target exploration

Which other drugs act on a target linked to your starting compound? Follow recorded drug and target relationships, compare mechanisms, and trace associated pathways and diseases to support drug discovery research.

DrugStarting pointTargetSharedDrug BSame targetDrug CSame targetacts onDrugStarting pointTargetSharedDrug BSame targetDrug CSame targetacts on
View Cypher example

Set $drugId to the starting drug. ACTS_ON carries a recorded mechanism; this example requires matching mechanism values.

MATCH (start:Drug)-[a:ACTS_ON]->(target:Target)
      <-[b:ACTS_ON]-(other:Drug)
WHERE start.id = $drugId AND other <> start
  AND a.mechanism = b.mechanism
RETURN DISTINCT other.id AS relatedDrug,
       target.id AS sharedTarget, b.mechanism AS mechanism
LIMIT 25

Returns: Other drugs sharing both a target and a recorded mechanism with the starting drug. These relationships support research review, not therapeutic equivalence.

02

Patient cohorts & treatment pathways

Which patients match the diagnosis and treatment criteria for a research cohort? Connect clinical records to identify matching patients, then trace their recorded treatments and disease progression for further analysis.

Initial diagnosisAge 50–70PatientMatches bothAdjuvanttreatmentRadiochemotherapyInitial diagnosisAge 50–70PatientMatches bothAdjuvanttreatmentRadiochemotherapy
View Cypher example

The Hancock oncology demo matches patients aged 50–70 at initial diagnosis with a recorded adjuvant radiochemotherapy treatment.

MATCH path = (p:Patient)-[:RECEIVED_INITIAL_DIAGNOSIS]->(d:InitialDiagnosis)
MATCH path2 = (p)-[:UNDERWENT_ADJUVANT_TREATMENT]->(a:AdjuvantTreatment)
WHERE d.age_at_initial_diagnosis >= 50
AND d.age_at_initial_diagnosis <= 70
AND a.adjuvant_radiochemotherapy = 'yes'
RETURN path, path2

Returns: Patient-to-diagnosis and patient-to-treatment paths for patients meeting both criteria.

03

Provider, pharmacy & prescription networks

Which other pharmacies fill prescriptions from the same prescriber? Follow prescription records to find the connected pharmacies, then inspect their patient and claims relationships as part of a pharmacy review.

PharmacyStarting pointPrescriberSharedPharmacy BAlso connectedPharmacy CAlso connecteddispensed forPharmacyStarting pointPrescriberSharedPharmacy BAlso connectedPharmacy CAlso connecteddispensed for
View Cypher example

Set $pharmacyId to the pharmacy under review. DISPENSED_FOR maps pharmacy/prescriber relationships prepared from prescription records.

MATCH (start:Pharmacy)-[:DISPENSED_FOR]->(prescriber:Prescriber)
      <-[:DISPENSED_FOR]-(other:Pharmacy)
WHERE start.id = $pharmacyId AND other <> start
RETURN DISTINCT other.id AS relatedPharmacy,
       prescriber.id AS sharedPrescriber
LIMIT 25

Returns: Other pharmacies and the prescribers they share with the starting pharmacy. Shared prescribers alone do not imply inappropriate activity.

02 / Graph analytics on clinical & research data

Run graph analytics on your existing tables.

Your tables already record drugs, targets, diagnoses, encounters, and prescriptions. Map those records and their relationships into a graph model, then follow the connections with PuppyGraph.

Traditional approach

Copy data. Keep two systems in sync.

ETL + duplicate storage + ongoing sync

  1. Curated clinical & research tables
    IcebergDatabricksBigQueryPostgreSQL
  2. Build and run a graph-loading pipeline
    ExportTransformLoad
    Export, transform, load, then synchronize changes
  3. Operate a separate graph databaseStore and maintain another copy of the data
    Keep the graph copy in sync
  4. Healthcare & life sciences analytics
  • More pipeline work. Propagate updates, backfills, and schema changes into the graph copy.
  • Another system to operate. Manage graph storage and the synchronization jobs alongside your data platform.

With PuppyGraph

Query existing data. Skip the graph copy.

Zero-ETL · Query data in place

  1. Curated clinical & research tables
    IcebergDatabricksBigQueryPostgreSQL
  2. PuppyGraph queries source tablesMap tables to vertices and edges in a graph schema
  3. Healthcare & life sciences analytics
  • No graph-loading pipeline. Query your existing clinical and research tables through the graph model.
  • No separate graph data store. Keep your source data in the platform you already operate.

03 / Pharmacy analysis in practice

Rank pharmacies by their share of flagged patients.

Follow each pharmacy’s claims to its distinct patients, then compare the share with an existing flag. This example adapts a real customer question using synthetic data.

filedfor patientPharmacy P1Starting pointClaimsGrouped for illustrationP-101FlaggedP-102FlaggedP-103FlaggedP-104Not flaggedfiledfor patientPharmacy P1Starting pointClaimsGrouped for illustrationP-101FlaggedP-102FlaggedP-103FlaggedP-104Not flagged

From claims to patient share

Count patients, then compare proportions.

  1. Follow each pharmacy’s claims to its patients, counting each patient once per pharmacy. P1 has four distinct patients.
  2. Count patients with an existing flag, such as a high-cost flag. P1 has three flagged patients.
  3. Divide flagged patients by total distinct patients, then rank pharmacies by that share. For P1, 3 ÷ 4 = 75%.
Flagged share by pharmacy, synthetic sample
PharmacyFlagged / total patientsShare
P13 / 475%
P23 / 1225%
P31 / 812.5%

P1 and P2 carry the same flagged count and rank differently once the total is included. Use the share and patient counts together to prioritize further review of the associated claims and patients.

Build your healthcare graph

Bring graph context to healthcare and life sciences workflows.

Download PuppyGraph to query your clinical and research data, book a demo for your use case, or join our Slack community to ask questions.

$ docker run -p 8081:8081 \
  -p 8182:8182 -p 7687:7687 \
  puppygraph/puppygraph:latestPulling puppygraph/puppygraph:latestStarting PuppyGraphWeb console listening on 0.0.0.0:8081Gremlin endpoint listening on 0.0.0.0:8182PuppyGraph is running# no pipelines · no second database · no data movement