EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Protein-ligand binding affinity prediction using multi-instance learning with docking structures

Hyojin Kim, Heesung Shim, Aditya Ranganath, Stewart He, Garrett Stevenson, Jonathan E Allen

Frontiers in Pharmacology · 03.01.2025 · 15

DOI: 10.3389/fphar.2024.1518875 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Pharmacology, online erschienen am 03.01.2025. Verfasst von Hyojin Kim und 5 weiteren Autor:innen. Beteiligte Einrichtung: Center for Applied Scientific Computing, Lawrence Livermore National Laboratory, Livermore, CA, United States. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 54 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Recent advances in 3D structure-based deep learning approaches demonstrate improved accuracy in predicting protein-ligand binding affinity in drug discovery. These methods complement physics-based computational modeling such as molecular docking for virtual high-throughput screening. Despite recent advances and improved predictive performance, most methods in this category primarily rely on utilizing co-crystal complex structures and experimentally measured binding affinities as both input and output data for model training. Nevertheless, co-crystal complex structures are not readily available and the inaccurate predicted structures from molecular docking can degrade the accuracy of the machine learning methods.

Methods

We introduce a novel structure-based inference method utilizing multiple molecular docking poses for each complex entity. Our proposed method employs multi-instance learning with an attention network to predict binding affinity from a collection of docking poses.

Results

We validate our method using multiple datasets, including PDBbind and compounds targeting the main protease of SARS-CoV-2. The results demonstrate that our method leveraging docking poses is competitive with other state-of-the-art inference models that depend on co-crystal structures.

Discussion

This method offers binding affinity prediction without requiring co-crystal structures, thereby increasing its applicability to protein targets lacking such data.

Schlagwörter

Stichwörter der Autor:innen

3D atomic graph representationAI-driven drug developmentattention mechanismmolecular dockingmulti-instance learningprotein-ligand interactionstructure-based machine learningvirtual high-throughput screening

Autor:innen und Institutionen

  • Center for Applied Scientific Computing, Lawrence Livermore National Laboratory, Livermore, CA, United States.
  • Biosciences and Biotechnology Division, Lawrence Livermore National Laboratory, Livermore, CA, United States.
  • Global Security Computing Applications Division, Lawrence Livermore National Laboratory, Livermore, CA, United States.
  • Computational Engineering Division, Lawrence Livermore National Laboratory, Livermore, CA, United States.

Zitieren

APA

Kim, H., Shim, H., Ranganath, A., He, S., Stevenson, G., & Allen, J. E. (2025). Protein-ligand binding affinity prediction using multi-instance learning with docking structures. Frontiers in Pharmacology, 15. https://doi.org/10.3389/fphar.2024.1518875

Vancouver

Kim H, Shim H, Ranganath A, He S, Stevenson G, Allen JE. Protein-ligand binding affinity prediction using multi-instance learning with docking structures. Frontiers in Pharmacology. 2025;15. doi:10.3389/fphar.2024.1518875

BibTeX
@article{Kim2025,
  author = {Kim, Hyojin and Shim, Heesung and Ranganath, Aditya and He, Stewart and Stevenson, Garrett and Allen, Jonathan E},
  title = {Protein-ligand binding affinity prediction using multi-instance learning with docking structures},
  journal = {Frontiers in Pharmacology},
  year = {2025},
  volume = {15},
  doi = {10.3389/fphar.2024.1518875}
}
Kostenlos starten →