GRIPHIN: grids of pharmacophore interaction fields for affinity prediction

2026/05/07 by Daniel M. Rose, Thomas Seidel, Thierry Langer · 1 voice
Biochemistry, Genetics and Molecular Biology · Computer Science · #Cell Image Analysis Techniques #Computational Drug Discovery Methods

paper · doi:10.1186/s13321-026-01203-8

openalex created_date 2025/10/10 · openalex publication_date 2026/05/07 · openalex updated_date 2026/07/29

Abstract

Pharmacophores are widely used to describe protein-ligand interactions. In this work, we propose a hybrid framework for binding affinity prediction that combines pharmacophoric maps of the protein binding site with a graph-based representation of the ligand. Our method achieves performance comparable to state-of-the-art models while offering interpretability through attribution methods, thereby demonstrating the potential of pharmacophoric representations in deep learning.Scientific contributionWe investigate whether a purely pharmacophoric representation of the protein pocket is sufficient to train a deep learning model for affinity prediction. For this purpose, we devise a hybrid model architecture from simple building blocks for affinity prediction. To enhance interpretability, we apply integrated gradients to attribute predictions to individual pharmacophoric features. Source code and model weights are available at https://github.com/molinfo-vienna/GRIPHIN .

Citations

• AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading
• RCSB Protein Data Bank (RCSB.org): delivery of experimentally-determined PDB structures alongside one million computed structure models of proteins from artificial intelligence/machine learning
• How Attentive are Graph Attention Networks?
• Captum: A unified and generic model interpretability library for PyTorch
• PIGNet: A physics-informed deep learning model toward generalized drug-target interaction predictions
• Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
• Improved Protein-ligand Binding Affinity Prediction with Structure-Based Deep Fusion Inference
• Fast Graph Representation Learning with PyTorch Geometric
• Graph Attention Networks
• Attention Is All You Need
• Axiomatic Attribution for Deep Networks
• Generalization in Adaptive Data Analysis and Holdout Reuse
• ImageNet classification with deep convolutional neural networks

Discussions

• the third paper in the #iccs2025 collection is out: https://doi.org/10.1186/s13321-026-01203-8 "GRIPHIN: grids of pharmacophore interaction fields for affinity prediction" "We investigate whether a [bsky, 0 points, 0 comments]

Related

• Binding Affinity Prediction: From Conventional to Machine Learning-Based Approaches
• A Point Cloud-Based Deep Learning Strategy for Protein-Ligand Binding Affinity Prediction
• PIGNet2: A Versatile Deep Learning-based Protein-Ligand Interaction Prediction Model for Binding Affinity Scoring and Virtual Screening
• DeepRLI: A Multi-objective Framework for Universal Protein--Ligand Interaction Prediction
• On Machine Learning Approaches for Protein-Ligand Binding Affinity Prediction