MOLECULAR SOLUBILITY PREDICTION WEB APP

This repo was mainly created for learning to create web apps using streamlit.

This tool predicts the molecular solubility of a compound given its SMILES string.

It also calculates shap values and plots some graphics to explain the models predictions.

It is possible to acess prediction from a Lineal Regression and a XGBoost models.

Requirements:

Clone the repository:

git clone https://github.com/jcorreia11/streamlit-solubility.git

Install the requirements:

pandas~=1.4.2
streamlit~=1.10.0
shap~=0.41.0
xgboost~=1.6.1
rdkit-pypi~=2022.3.3
streamlit_shap~=1.0.2

If you want to experiment with the solubility_models.ipynb notebook you also need:

scikit-learn~=1.0.1
matplotlib~=3.5.1

Run the tool in your browser:

Run the following command in your terminal in the solubility_app.py directory:

streamlit run solubility_app.py 

Paste your SMILES strings in the SMILES input box and explore!

Deploy (TODO)

1. Heroku
2. Streamlit Sharing