EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Phase‐Resolved Geometric Deep Learning Framework Maps Structural Determinants of Disease‐Associated Protein Aggregation and Guides Suppressor Design

Jia Shen Sio iD, Wei Xuan Wilson Loo iD, Yan Shan Loo iD, Wen Xin Tan, Hui Xuan Lim iD, Huitao Liu iD, Chen Seng Ng iD

Advanced Science · 2026

DOI: 10.1002/advs.76118 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advanced Science, 2026. Verfasst von Jia Shen Sio und 6 weiteren Autor:innen. Beteiligte Einrichtung: School of Science, Monash University Malaysia, Jalan Lagoon Selatan, Selangor, Malaysia. Hauptthemen laut MEDLINE: Deep Learning, Protein Aggregation, Pathological, Protein Aggregates, Neurodegenerative Diseases. Die Arbeit nennt 65 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: IBRO Early Career Principal Investigator grant, Sunway University, Henan University of Technology.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Protein aggregation drives major neurodegenerative diseases, yet most computational predictors collapse assembly into static risk scores and do not resolve the distinct structural determinants of nucleation and elongation. Here, we present SKALE 2.0, a phase‐resolved geometric deep learning framework that represents proteins as multimodal structural graphs and learns mutation‐induced aggregation phenotypes directly from three‐dimensional topology. Across SOD1, TDP‐43, MAPT, and PRNP, SKALE 2.0 recovered a conserved latent transition from nucleation to elongation while resolving distinct mutation‐specific phase sensitivities. Representative protein language model, AlphaFold‐derived feature, and non‐phase‐aware structural baselines failed to recover both phase‐dependent mutation modulation and phase separability, indicating that explicit phase conditioning is essential. The learned geometry showed that nucleation is preferentially coupled to buried hydrophobic perturbations, whereas elongation is shaped by solvent‐accessible interfaces that support fibril propagation. This framework explains how pathogenic variants can remain globally folded yet acquire aggregation competence through localized structural rewiring. Recombinant SOD1 experiments validated predicted suppressor, enhancer, and phase‐switch mutations, demonstrating that initiation and propagation can be tuned independently. SKALE 2.0 links atomic topology to phase‐specific assembly kinetics and enables a constraint‐aware design of aggregation suppressors.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningProtein Aggregation, PathologicalProtein AggregatesNeurodegenerative DiseasesHumansSuperoxide Dismutase-1MutationDNA-Binding Proteinstau Proteins

Stichwörter der Autor:innen

artificial intelligencecomputational biologydeep learningenhancerfibrillanguage modelnucleationprnpprotein aggregationsuppressor

Substanzen

Superoxide Dismutase-1Protein AggregatesDNA-Binding Proteinstau ProteinsSOD1 protein, human

Autor:innen und Institutionen

  • School of Science, Monash University Malaysia, Jalan Lagoon Selatan, Selangor, Malaysia.
  • Department of Pharmaceutics and Biopharmaceutics, University of Marburg, Robert-Koch-Str. 4, Marburg, Germany.
  • Department of Biomedical Sciences, Sir Jeffrey Cheah Sunway Medical School, Faculty of Medical and Life Sciences, Sunway University, Jalan Universiti, Selangor, Malaysia.
  • Sunway Microbiome Centre, Faculty of Medical and Life Sciences, Sunway University, Jalan Universiti, Selangor, Malaysia.
  • Henan Linker Technology Key Laboratory, College of Advanced Interdisciplinary Science and Technology, Henan University of Technology, Zhengzhou, China.

ORCID-Profile: Jia Shen Sio, Wei Xuan Wilson Loo, Yan Shan Loo, Hui Xuan Lim, Huitao Liu, Chen Seng Ng

Förderung

  • IBRO Early Career Principal Investigator grant
  • Sunway University
  • Henan University of Technology
  • Shenzhen Vaccine Biopharmaceuticals Limited
  • Alexander von Humboldt Foundation
  • Alexander von Humboldt-Stiftung

Zitieren

APA

Sio, J. S., Loo, W. X. W., Loo, Y. S., Tan, W. X., Lim, H. X., Liu, H., & Ng, C. S. (2026). A Phase‐Resolved Geometric Deep Learning Framework Maps Structural Determinants of Disease‐Associated Protein Aggregation and Guides Suppressor Design. Advanced Science. https://doi.org/10.1002/advs.76118

Vancouver

Sio JS, Loo WXW, Loo YS, Tan WX, Lim HX, Liu H, et al. A Phase‐Resolved Geometric Deep Learning Framework Maps Structural Determinants of Disease‐Associated Protein Aggregation and Guides Suppressor Design. Advanced Science. 2026. doi:10.1002/advs.76118

BibTeX
@article{Sio2026,
  author = {Sio, Jia Shen and Loo, Wei Xuan Wilson and Loo, Yan Shan and Tan, Wen Xin and Lim, Hui Xuan and Liu, Huitao and Ng, Chen Seng},
  title = {A Phase‐Resolved Geometric Deep Learning Framework Maps Structural Determinants of Disease‐Associated Protein Aggregation and Guides Suppressor Design},
  journal = {Advanced Science},
  year = {2026},
  doi = {10.1002/advs.76118}
}
Kostenlos starten →