Architecture

Backend Architecture

• Written in Python using the FastAPI framework.
• Provides RESTful APIs for cheminformatics operations.
• Hybrid toolkit integration:
  • Native Python toolkits: RDKit, Open Babel.
  • Java toolkits (e.g., CDK, SRU, OPSIN) accessed via JPype.
• Docker containerization ensures reproducibility and consistent deployment.
• Designed for scalability and future extensibility without disrupting current services.
• Publicly available at https://api.naturalproducts.net/latest/docs.

Frontend Architecture

• Built using React for strong community support and sustainability.
• Component-based design with modular service layers and custom hooks.
• Styled using Tailwind CSS for responsive UI.
• Communicates with backend via RESTful API using Axios.
• API interactions are abstracted through a dedicated service layer.
• Structured into functionally specialized pages.
• Publicly available at https://app.naturalproducts.net/.

Toolkit Integration Challenge

• Toolkits use different languages:
  • RDKit/OpenBabel: C++/Python
  • CDK: Java
• Integrating multiple toolkits is complex due to compatibility and environment setup.
• Solution: use containerization and microservices for abstraction and flexibility.

Microservices & Containerization

• Microservices: each function is a separate, scalable, maintainable unit.
• Containers provide isolated, reproducible environments.
• Docker is used for packaging and sharing toolkits.
• CM (Cheminformatics Microservice) is publicly available via Docker Hub.

API Design

• Uses REST API based on OpenAPI Specification 3.1.0.
• Built with FastAPI for speed, efficiency, and automatic documentation.
• Promotes interoperability, validation, and tool integration.

Functional Scope

• Provides:
  • Format conversions
  • Optical Structure Recognition (OSR)
  • Chemical data standardization
  • Descriptor calculations
• Integrated tools:
  • Cheminformatics: RDKit, CDK, OpenBabel
  • Deep learning: DECIMER, STOUT
• Enables scalable, interoperable, and efficient cheminformatics applications.

Architectural Decisions

• Combines FastAPI + Docker for seamless deployment.
• Maintains flexibility via microservice architecture:
  • Individual services can be updated without affecting others.
• Contrary to usual practice, all toolkits are packaged in one container:
  • Simplifies deployment and avoids complex orchestration.

Deployment & Monitoring

• Dockerfile, docker-compose YAML, and deployment scripts are available on GitHub.
• HELM charts provided for Kubernetes-based deployments.
• Monitoring and visualization:
  • Prometheus for metrics collection and alerting.
  • Grafana for real-time dashboards and usage visualization.