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

Meta-Learning as a Promising Strategy for Lipid Nanoparticle Optimization and Ionizable Lipid Discovery

Felix Sieber-Schäfer, Lasse Hagedorn, Leon Reger, Katharina Möbius, Benjamin Winkeljann iD, Olivia M Merkel iD

Nano Letters · 03.06.2026 · 26 (21) · 6928-6935

DOI: 10.1021/acs.nanolett.6c00782 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Nano Letters, online erschienen am 21.05.2026. Verfasst von Felix Sieber-Schäfer und 5 weiteren Autor:innen. Beteiligte Einrichtung: Ludwig-Maximilians-Universität München, Department of Pharmacy, Butenandtstraße 5, 81377 Munich, Germany. Hauptthemen laut MEDLINE: Lipids, Machine Learning, Nanoparticles, RNA, Small Interfering. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: H2020 European Research Council.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The rapid growth of lipid nanoparticle (LNP)-based RNA therapeutics demands predictive tools to accelerate formulation and lipid design, yet development remains limited by complex delivery mechanisms and scarce high-quality data. We investigated few-shot meta-learning (FSL) as a strategy for early stage, data-limited LNP development using a published data set. Meta-learning tasks were constructed from data provenance and formulation conditions, and several FSL methods were benchmarked against supervised baselines using fingerprint- and graph-based representations. In a stringent extrapolation setting, all siRNA data were excluded from meta-training and reserved for testing. Under this protocol, model-agnostic meta-learning (MAML) outperformed both supervised and transfer-learning baselines, achieving an average R2 of 0.38 ± 0.049 on the siRNA holdout task, whereas non-meta-learning models performed near zero. In retrospective active-learning simulations and experimental validation with 15 newly synthesized ionizable lipids across multiple cell lines and RNA cargos, MAML consistently surpassed random forest (RF), supporting FSL as a promising framework for data-scarce RNA delivery design tasks.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

LipidsMachine LearningNanoparticlesRNA, Small InterferingHumansRandom ForestLiposomes

Stichwörter der Autor:innen

Few-Shot LearningLipid NanoparticleLipidsMachine LearningMeta-Learning

Substanzen

Lipid NanoparticlesLipidsRNA, Small InterferingLiposomes

Autor:innen und Institutionen

  • Ludwig-Maximilians-Universität München, Department of Pharmacy, Butenandtstraße 5, 81377 Munich, Germany.
  • Member of the German Center for Lung Research (DZL), 81377 Munich, Germany.
  • Soft Matter Chemistry, Department of Chemistry, Faculty of Science, University of Helsinki, Helsinki 00014, Finland.
  • Center for Nanoscience (CeNS), 80799 Munich, Germany.

ORCID-Profile: Benjamin Winkeljann, Olivia M Merkel

Förderung

  • H2020 European Research Council

Zitieren

APA

Sieber-Schäfer, F., Hagedorn, L., Reger, L., Möbius, K., Winkeljann, B., & Merkel, O. M. (2026). Meta-Learning as a Promising Strategy for Lipid Nanoparticle Optimization and Ionizable Lipid Discovery. Nano Letters, 26(21), 6928-6935. https://doi.org/10.1021/acs.nanolett.6c00782

Vancouver

Sieber-Schäfer F, Hagedorn L, Reger L, Möbius K, Winkeljann B, Merkel OM. Meta-Learning as a Promising Strategy for Lipid Nanoparticle Optimization and Ionizable Lipid Discovery. Nano Letters. 2026;26(21):6928-6935. doi:10.1021/acs.nanolett.6c00782

BibTeX
@article{SieberSchfer2026,
  author = {Sieber-Schäfer, Felix and Hagedorn, Lasse and Reger, Leon and Möbius, Katharina and Winkeljann, Benjamin and Merkel, Olivia M},
  title = {Meta-Learning as a Promising Strategy for Lipid Nanoparticle Optimization and Ionizable Lipid Discovery},
  journal = {Nano Letters},
  year = {2026},
  volume = {26},
  number = {21},
  pages = {6928--6935},
  doi = {10.1021/acs.nanolett.6c00782}
}
Kostenlos starten →