snapms

core code for the SNAP MS platform for predicting identities of natural products from MS data

Installation

NEW - uses Poetry Python. To setup local dev simply run poetry install

Requires Python 3.8+.

The docker-compose solution has been updated such that it should work with the following:

docker-compose build
docker-compose up -d

Django App

This repo includes an example Django App for Snap MS. Some configuration is required.

Environment file (.env)

REDIS_URI=redis://localhost:6379/0
CYTOSCAPE_DATADIR=/root/data
SNAPMS_DATADIR=/home/username/git/snapms/data
NPATLAS_FILE=/home/username/git/snapms/data/atlas_input/NPAtlas_download.json
COCONUT_FILE=/home/username/git/snapms/data/atlas_input/COCONUT_download.json

The SNAPMS_DATADIR MUST exist already and the NPATLAS_FILE and COCONUT_FILE MUST also be available.

To run locally you must also create a DB directory 'db' as 'snapms/db'

Running the development server requires two instances. For each instance, open a terminal window, navigate to the root 'snapms' directory and type:

1. Django App

dotenv run ./manage.py runserver

2. RQ worker

dotenv run ./manage.py rqworker high default low

You must also have the following two Docker containers running locally:

Requirements

1. Redis - Docker

You can easily start Redis in a docker container locally with

docker run -itd -p 6379:6379 --name snapms-redis redis

2. Cytoscape

Cytoscape automation requires a running instance. You can run it on the desktop, but on a server it will be easier to run it in a docker container. The CyREST runs on port 1234.

For cytoscape session file serving to work, the mounted volume should point to the same place as the SNAPMS_DATADIR.

In development, using ./data as the SNAPMS_DATADIR is sensible.

docker run --name cy -itd -v $(pwd)/data/testing:/root/data -p 1234:1234 jvansan/cytoscape-desktop-headless:latest

Docker deployment

In this repo is a docker-compose.yml file which shows how to deploy this example app using Traefik as a reverse proxy. In a real production environment, you should NOT use the build in Django runserver.