Antioxidant Activity: In silico models for predicting the antioxidant activity of small molecules relevant to human health

Our software has been developed for predicting the antioxidant activity of small molecules (< 1000 Da) and it aims to assist in identifying potential substances that could be applied in health support. It is built on regression models developed on an expert-curated dataset of antioxidants. Given the SMILES as input the software will predict the half-maximal inhibitory concentration (IC50) of the substance(s) of interest.

Installation

1. Download from here the AntioxidantActivity_DPPH folder and unzip it
2. Do not move the files out of the folder.
3. Install Anaconda prompt following the instruction here
4. Open Anaconda terminal and create the AntioxidantDPPH environment with the command: > conda create --name AntioxidantActivity_DPPH python=3.11
5. Activate your environment with the command:

6.      > conda activate AntioxidantActivity_DPPH
   

7. Install the following dependencies with the command:

8.      > pip install scikit-learn==1.4.0 rdkit==2023.9.4 pandas==2.2.0 mordred==1.2.0 xgboost==2.1.3
   

9. Move to the AntioxidantActivity_DPPH folder before prompt the target compound

Usage

To run the program:

Single molecule mode:

Command -> python Main.py --smiles [write single SMILES] [optional]: --summary 1

Batch mode:

Command -> python Main.py--file [add file name] [optional]: --summary 1 Key: --file: path of file to predict the antioxidant activity must have column named SMILES (batch functionality)

OPTIONAL: set summary to one

    --summary 1 to obtain only the consensus prediction and uncertanty value.
    [default] --summary None to obtain all the models' predictions.

Examples:

python Main.py --smiles c1ccccc1CCN --summary 1
python Main.py --file test.xlsx --summary 1 
python Main.py --file test.xlsx