EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Protein Binding Site Representation in Latent Space

Frederieke Lohmann iD, Stephan Allenspach iD, Kenneth Atz iD, Carl C G Schiebroek iD, Jan A Hiss iD, Gisbert Schneider iD

Molecular Informatics · 2025

DOI: 10.1002/minf.202400205 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Molecular Informatics, 2025. Verfasst von Frederieke Lohmann und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Chemistry and Applied Biosciences, ETH Zurich, Vladimir-Prelog-Weg 4, 8093, Zürich, Switzerland. Hauptthemen laut MEDLINE: Proteins. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Interpretability and reliability of deep learning models are important for computer‐based drug discovery. Aiming to understand feature perception by such a model, we investigate a graph neural network for affinity prediction of protein‐ligand complexes. We assess a latent representation of ligand binding sites and investigate underlying geometric structure in this latent space and its relation to protein function. We introduce an automated computational pipeline for dimensionality reduction, clustering, hypothesis testing, and visualization of latent space. The results indicate that the learned protein latent space is inherently structured and not randomly distributed. Several of the identified protein binding site clusters in latent space correspond to functional protein families. Ligand size was found to be a determinant of cluster geometry. The computational pipeline proved applicable to latent space analysis and interpretation and can be adapted to work for different datasets and deep learning models.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

ProteinsBinding SitesDeep LearningLigandsNeural Networks, ComputerProtein BindingDrug Discovery

Stichwörter der Autor:innen

drug discoveryinterpretabilitymachine learningneural networkprotein structure

Substanzen

ProteinsLigands

Autor:innen und Institutionen

  • Department of Chemistry and Applied Biosciences, ETH Zurich, Vladimir-Prelog-Weg 4, 8093, Zürich, Switzerland.
  • Department of Biosystems Science and Engineering, ETH Zurich, Klingelbergstrasse 48, 4056, Basel, Switzerland.

ORCID-Profile: Frederieke Lohmann, Stephan Allenspach, Kenneth Atz, Carl C G Schiebroek, Jan A Hiss, Gisbert Schneider

Zitieren

APA

Lohmann, F., Allenspach, S., Atz, K., Schiebroek, C. C. G., Hiss, J. A., & Schneider, G. (2025). Protein Binding Site Representation in Latent Space. Molecular Informatics. https://doi.org/10.1002/minf.202400205

Vancouver

Lohmann F, Allenspach S, Atz K, Schiebroek CCG, Hiss JA, Schneider G. Protein Binding Site Representation in Latent Space. Molecular Informatics. 2025. doi:10.1002/minf.202400205

BibTeX
@article{Lohmann2025,
  author = {Lohmann, Frederieke and Allenspach, Stephan and Atz, Kenneth and Schiebroek, Carl C G and Hiss, Jan A and Schneider, Gisbert},
  title = {Protein Binding Site Representation in Latent Space},
  journal = {Molecular Informatics},
  year = {2025},
  doi = {10.1002/minf.202400205}
}
Kostenlos starten →