CTS-Lite

A lightweight Chemical Translation Service using a curated subset of 13 million compounds from the PubChem database

Project Health

Website

• https://cts-lite.metabolomics.us/

API Usage

• Please refer to the documentation for information regarding the API

Credits

• PubChem - database used for compound data
• RDKit - library used for SMILES to InChIKey conversion
• Wishart Research Group - Chemical Taxonomy using ClassyFire

Development

Stack

• Go
• SQLite
• JavaScript/HTML/CSS
• Docker
• Playwright
• Locust (load testing)

Testing

• Run unit tests with go test ./...
• Run E2E playwright tests with cd playwright && npm test

Docker

• CTS-Lite is containerized with Docker
• The GitHub Actions workflow will automatically build and deploy the complete application image upon any push or merge to the main branch
• Note: to build the docker image locally, you must have the database built and stored as dataset/compounds.db
  • cd dataset && go run cmd/build-db/build-db.go cts-lite.csv compounds.db

Dataset Creation

• To create the csv dataset, simply run the create_csv_dataset.sh script found under cmd/
• To update the dataset used by production, make sure you elect to push to S3 at the end of the script
  • Then, the next time the app is deployed via GitHub Actions (push/merge to main), the latest dataset will be downloaded from S3 and the database will be rebuilt
• To create a local instance of compounds.db (SQLite database used by the app), run the build-db module like so:
  • cd dataset && go run cmd/build-db/build-db.go cts-lite.csv compounds.db