# Improving detection of protein-ligand binding sites with 3D segmentation

## TL;DR

This paper presents a 3D convolutional neural network that improves the detection of druggable binding pockets on protein surfaces, aiding early drug discovery with high accuracy and interpretability.

## Contribution

Developed a novel 3D fully convolutional neural network for protein binding site segmentation, enhancing prediction accuracy and usability in drug discovery workflows.

## Key findings

- High prediction accuracy for binding site segmentation
- Provides intuitive visual representations of binding pockets
- Code and tools are openly available for researchers

## Abstract

In recent years machine learning (ML) took bio- and cheminformatics fields by storm, providing new solutions for a vast repertoire of problems related to protein sequence, structure, and interactions analysis. ML techniques, deep neural networks especially, were proven more effective than classical models for tasks like predicting binding affinity for molecular complex. In this work we investigated the earlier stage of drug discovery process - finding druggable pockets on protein surface, that can be later used to design active molecules. For this purpose we developed a 3D fully convolutional neural network capable of binding site segmentation. Our solution has high prediction accuracy and provides intuitive representations of the results, which makes it easy to incorporate into drug discovery projects. The model's source code, together with scripts for most common use-cases is freely available at http://gitlab.com/cheminfIBB/kalasanty