README.MD

Getting Started

Python implementation of the main clustering results for the paper entitled: On the best way to cluster NCI-60 molecules

Prerequisites

• Python 3.9.7
• venv

Installation

1. Create the virtual environment

python -m venv .venv

2. Activate the virtual environment
   1. On Windows, run:

   .venv\Scripts\activate.bat

   2. On Linux or MacOs, run:

   .source .venv/bin/activate

3. Install all the necessary packages

python -m pip install -r requirements.txt

Usage

1. Download and store the from the NCI-60 Growth Inhibition Data.

   1. The required files contain the endpoints calculated from concentration curves ("CANCER60GI50_Oct2020.LST", for instance) and SMILES ("Chem2D_Jun2016.smi", for instance). Other releases of both files are also available for download.

2. In the file main.py, replace the following lines.

   1. Replace with the directory where the downloaded data is stored

   dir_working = '/home/hernadez/Documents/NCI60_data/' 

   2. Replace only if the files are different from those suggested in 1.

   file_nci60 = 'CANCER60GI50_Oct2020.LST'  
   file_smiles = 'Chem2D_Jun2016.smi' 

3. By default, the number of clusters (k) has been set to 7, and the removal of outliers has been requested. Both values can be modified in the following lines.

k = 7  # number of clusters
outliers = True  # to remove outliers

To run:

python main.py

Output: As a result, a folder containing the clustering assignment ([NSC, SMILES, Cluster ID]) and the corresponding clustering quality metrics will be created.

Contact

1. Saiveth Hernández-Hernández, email: saiveth.hernadez@inserm.fr
2. Pedro J.Ballester, email: p.ballester@imperial.ac.uk