Uncertainty in Binding Affinity: A Deep Learning Benchmark

Citation

Cite this Uncertainty quantification enables reliable deep learning for protein–ligand binding affinity prediction.

Contact

Milad Rayka, milad.rayka@yahoo.com

Install

1- First install python (3.8.16) then make a virtual environment and activate it.

python -m venv env
.\env\Scripts\activate

Which env is the location to create the virtual environment.

2- Clone uncertainty_quantification Github repository.

git clone https://github.com/miladrayka/uncertainty_quantification.git

3- Change your directory to uncertainty_quantification.

4- Install required packages with pip.

pip install -r requirements.txt

Usage

To reproduce all results, tables, and figures, refer to analysis folder.

FFNN-ECIF-BayesByBackprop GUI

FFNN-ECIF-BayeByBackprop GUI (Graphical User Interface) is a software for protein-ligand binding affinity prediction and uncertainty quantification. The model is based on FeedForward-NeuralNetwork (FFNN) and Extended-Connectivity Interaction Feature (ECIF) for binding affinity prediction. We augment our model with Bayes by the Backprop approach for uncertainty quantification of predicted binding affinity. See software folder for more information.

Copy Right

Copyright (c) 2025, Milad Rayka