CDK9 Screen

CDK9 Screen tool allows you to predict the potential inhibitory activity of small molecules against human CDK9. This repository provides necessary Python scripts to perform predictions using a predefined environment and dependencies.

Citation

If you use CDK9Screen, please cite:

Piazza, L.; Poles, C.; Bononi, G.; Granchi, C.; Di Stefano, M.; Poli, G.; Giordano, A.; Medugno, A.; Napolitano, G.M.; Tuccinardi, T.; Alfano, L. Machine Learning-Based Virtual Screening for the Identification of Novel CDK-9 Inhibitors. Biomolecules 2026, 16, 12. https://doi.org/10.3390/biom16010012

Installation

To set up the required environment, use the provided YAML file:

conda env create -f cdk9_env.yml

Then, activate the environment:

conda activate cdk9_env

Usage

Run the predictions with Python, specifying the required inputs:

python run.py -in input.csv -out output.csv

Input

• The input file must be in CSV format and contain a column named SMILES containing the molecular structures.
• An example input CSV file is included in the repository for testing purposes.

Output

• The run.py script will generate a CSV file containing an additional column named Out_predictions storing the predicted class.
• If a molecule is assigned to class 1, it is predicted to be active; conversely, class 0 corresponds to an "inactive" prediction.