This repository contains ML models for predicting Anti-SARS-CoV-2 Activities (Live Virus Infectivity, Viral Entry, Viral Replication) and for performing Similarity Search

• Website: http://drugcentral.org/Redial

• Requirements
• Installation and Initialization
  • Docker
    • Run Docker Images
    • Build your own images
  • Manual Start
• Docker API Content
• Contributors
• Funding
• Acknowledgements
• License

At the moment, a standard machine with CPUs will work.

Currently, we suggest two ways to run REDIAL-2020 locally: Docker and manually running from source. With Docker, you can either pull an image from DockerHub, or build one on your own.

To install Docker, just follow the docker documentation.

The latest Redial-2020 images are available on the Docker Hub.

• docker pull sirimullalab/redial-2020:latest
• docker run -p 5000:5000 sirimullalab/redial-2020:latest
• curl -F smiles='CCCC' localhost:5000/predict OR curl -d 'smiles=CCCC' localhost:5000/predict

1. git clone this repo
2. cd /path/to/this/repo
3. docker build -t redial-2020:latest -f Dockerfile .
4. docker run -p 5000:5000 redial-2020:latest
5. curl -F smiles='CCCC' localhost:5000/predict OR curl -d 'smiles=CCCC' localhost:5000/predict

NOTE: The API accepts form-request / post-request.

1. Install Miniconda, for your operating system, from https://conda.io/miniconda.html
2. git clone this repo
3. cd /path/to/this/repo
4. conda env create -f environment-redial-2020.yml
5. conda activate redial-2020 (or source activate redial-2020 for older versions of conda)
6. Example Commands :

python3 redial_batch_screen.py --input_path <PATH_TO_CSV_FILE> --output_path <PATH_TO_SAVE_JSON_FILE> --calculate_alogps <Bool>

python3 redial_batch_screen.py --input_path test.csv --output_path ./ --calculate_alogps 1

Note:

1. Input csv file should contain atleast one column having SMILES as column name.
2. Predictions along with ALOPGS calculations may be slow, due to OCHEM API limitations. For faster predictions, use --calculate_alogps 0

API URL: hostname:port/predict

• Found here - ./docker_api_content.yml

Srijan Verma, Govinda KC, Giovanni Bocci, Suman Sirimulla, Tudor I. Oprea

Suman Sirimulla acknowledge support from the National Science Foundation with NSF-PREM grant #DMR-1827745.

MIT License