Parthenon User Manual
Welcome to Parthenon — a next-generation unified outcomes research platform built on the OMOP Common Data Model v5.4. Parthenon replaces legacy OHDSI Atlas with a modern, performant, and extensible interface while remaining fully compatible with the OHDSI analytical toolchain including HADES, CohortGenerator, and Circe.
What is Parthenon?
Parthenon provides a single browser-based interface for the entire real-world evidence (RWE) research lifecycle. Starting from vocabulary exploration and concept set construction, researchers build patient cohorts using a visual builder, then run a full spectrum of analyses — characterization, incidence rates, treatment pathways, population-level estimation, patient-level prediction, self-controlled case series, and evidence synthesis — all without leaving the platform.
Beyond traditional OHDSI analytics, Parthenon extends into domains that Atlas never reached:
- Genomics — Upload VCF files, annotate variants against ClinVar, browse mutations in an interactive variant browser, and convene virtual tumor boards with AI-assisted interpretation.
- Medical Imaging — View DICOM studies with a built-in Cornerstone3D viewer, connect to PACS systems via WADO-RS, and incorporate imaging criteria into cohort definitions.
- Health Economics & Outcomes Research (HEOR) — Model cost-effectiveness, identify care gaps across populations, and run population-level economic analytics.
- FHIR R4 Integration — Connect to EHR systems using SMART Backend Services for automated bulk export and incremental sync of clinical data into your OMOP CDM.
- AI-Assisted Analysis — An integrated AI service powered by Ollama and MedGemma provides semantic concept search, natural-language cohort suggestions, clinical result interpretation, and genomic variant summarization.
Parthenon exposes a REST API compatible with OHDSI WebAPI, so existing tools, phenotype libraries, and study packages continue to work without modification.
Manual Structure
| Part | Chapters | Topic |
|---|---|---|
| I — Getting Started | 1--2 | Platform introduction, architecture, data sources |
| II — Vocabulary | 3--4 | Vocabulary browser, concept sets |
| III — Cohorts | 5--8 | Cohort expressions, building, generating, and management |
| IV — Analyses | 9--14 | All seven analysis types and study packages |
| V — Data Ingestion | 15--17 | Uploading data, schema mapping, concept mapping |
| VI — Data Explorer | 18--20 | Achilles characterization, data quality, population stats |
| VII — Patient Profiles | 21 | Individual patient timelines |
| VIII — Administration | 22--26 | Users, roles, auth providers, system configuration, audit |
| Migration Guide | -- | Migrating from Atlas to Parthenon |
| Appendices | A--G | Reference material |
Platform Architecture
Parthenon is a containerized multi-service application orchestrated with Docker Compose. Each service is purpose-built and independently scalable.
Service Summary
| Service | Technology | Purpose |
|---|---|---|
| Frontend | React 19, TypeScript, Vite 7, TailwindCSS v4 | Single-page application with dark crimson and gold theme |
| Backend API | Laravel 11, PHP 8.4, Sanctum | RESTful API, authentication, authorization, job dispatch |
| Queue Worker | Laravel Horizon, Redis 7 | Background jobs: cohort generation, Achilles, bulk import |
| AI Service | Python FastAPI, Ollama, MedGemma | Semantic search, NLP cohort suggestions, variant interpretation |
| R Runtime | R Plumber, HADES packages | Statistical analyses via JDBC to OMOP CDM |
| Database (App) | PostgreSQL 16 | Application metadata: users, sources, cohort definitions |
| Database (CDM) | PostgreSQL 17 | OMOP CDM clinical data, vocabularies, Achilles results |
| Cache | Redis 7 | Session storage, query cache, queue broker |
Platform Modules
Parthenon organizes functionality into the following modules:
Clinical Research
- Data Sources — Configure connections to one or more OMOP CDM databases with per-schema daimons.
- Vocabulary — Search 7.2M+ OMOP concepts with text and AI-powered semantic search; compare concepts side-by-side.
- Concept Sets — Build reusable concept lists with descendant, mapped, and exclusion flags.
- Cohort Builder — Visually define cohorts using inclusion criteria, censoring events, and temporal logic.
- Analyses — Seven analysis types: Characterization, Incidence Rates, Treatment Pathways, Population-Level Estimation (PLE), Patient-Level Prediction (PLP), Self-Controlled Case Series (SCCS), and Evidence Synthesis.
- Studies — Package multiple analyses into reproducible study definitions for multi-site execution.
- Patient Profiles — Drill into individual patient timelines with domain-specific event visualization.
Data Management
- Data Explorer — Achilles-powered characterization dashboards, data quality checks (DQD), and population statistics.
- Data Ingestion — Upload CSV/TSV files, map schemas to OMOP CDM tables, and map source codes to standard concepts.
- Vocabulary Management — Upload Athena vocabulary ZIP bundles to refresh concept tables (admin).
Advanced Modules
- Genomics — Upload VCF files, annotate against ClinVar, browse variants with an interactive variant browser, and run virtual tumor boards with AI-assisted interpretation.
- Imaging — View DICOM studies inline with Cornerstone3D, connect to PACS via WADO-RS/DICOMweb, and use imaging criteria in cohort definitions.
- HEOR — Health economics modeling, cost-effectiveness analysis, care gap identification, and population-level economic analytics.
Integration & Administration
- FHIR R4 — SMART Backend Services for automated bulk data export and incremental sync from EHR systems.
- Jobs — Monitor background tasks: cohort generation, Achilles runs, bulk imports, FHIR syncs.
- Administration — User management, role and permission assignment, authentication providers (SAML 2.0, OIDC), AI provider configuration, system health monitoring, FHIR connection management, and sync dashboard.
Research Workflow
The typical research workflow in Parthenon follows a structured pipeline from vocabulary exploration to publishable results:
If you are new to the OMOP Common Data Model, start with the Vocabulary Browser chapter to understand how clinical concepts are organized. The vocabulary is the foundation of every cohort definition and analysis in Parthenon.