MAIP model

The MAIP model is a machine-learning model for scoring small molecules by their predicted likelihood of blood-stage antimalarial activity. This Docker package takes compound identifiers and SMILES strings as input, standardises valid molecules, and writes a model_score that can be used to prioritise compounds for follow-up. Check details in the reference articles (CITATION.cff).

Input

Provide a CSV file with these columns:

id,smiles
compound1,CCO
compound2,c1ccccc1

The output is a predictions.csv file containing:

• id
• smiles
• standard_smiles
• model_score

How To

Build the Docker image:

docker build --platform linux/amd64 -t maip-model:latest .

Run predictions for a CSV file:

mkdir -p data
cp MAIP_example.csv data/

docker run --rm --platform linux/amd64 \
  -v "$PWD/data":/data \
  maip-model:latest /data/MAIP_example.csv /data/output

The results will be written to:

data/output/predictions.csv

If no output directory is provided, the image writes to an output directory next to the input CSV:

docker run --rm --platform linux/amd64 \
  -v "$PWD/data":/data \
  maip-model:latest /data/MAIP_example.csv

With Rancher Desktop, if Docker cannot find the daemon, point the CLI at Rancher's socket:

DOCKER_HOST=unix://$HOME/.rd/docker.sock docker build --platform linux/amd64 -t maip-model:latest .

License

This project is licensed under the MIT License. See LICENSE.

Citation

If you use MAIP in your work, please cite:

Bosc N, Felix E, Gardner JMF, et al. MAIP: An Open-Source Tool to Enrich High-Throughput Screening Output and Identify Novel, Druglike Molecules with Antimalarial Activity. ACS Medicinal Chemistry Letters. 2023;14(12):1733–1741. DOI: 10.1021/acsmedchemlett.3c00369

A machine-readable citation is provided in CITATION.cff.