A machine learning benchmarking framework for lipid nanoparticle transfection efficiency prediction
4 affiliations
- 1Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, USA
- 2Department of Engineering, University of Cambridge, Cambridge, UK
- 3Pharmaceutical Sciences Division, School of Pharmacy, University of Wisconsin-Madison, Madison, WI, USA
- 4Department of Pharmaceutical Sciences, Gatton College of Pharmacy, East Tennessee State University, Johnson City, TN, USA
Abstract
The discovery of new ionizable lipids for efficient lipid nanoparticle (LNP)-mediated RNA delivery remains a major bottleneck in RNA therapeutics development. Recent advances demonstrate the potential of machine learning (ML) models to predict transfection efficiency directly from lipid structure, enabling high-throughput virtual screening and accelerating lead identification. However, as new models for LNP transfection prediction continue to emerge, the lack of rigorous and standardized benchmarking poses a significant risk and may undermine confidence in their reliability for discovery. Here, we present a robust ML benchmarking framework for evaluating transfection prediction models based on ionizable lipid structures. This framework systematically benchmarks diverse molecular representations—including Morgan fingerprints, Expert RDKit descriptors, and Grover graph-based embeddings—paired with a broad range of ML architectures spanning traditional models, feedforward neural networks, and state-of-the-art graph-based methods. In addition, the presented framework supports assessment of model generalization through Murcko scaffold splitting and evaluates prediction reliability beyond standard regression metrics by incorporating analyses of relative error distributions and ranking accuracy. Using a curated dataset of 1100 unique ionizable lipid structures derived from the HeLa transfection dataset originally reported by Xu et al. 1 , we show that within this framework, models leveraging explicit molecular substructure encoding—particularly multilayer perceptrons (MLPs) trained on Morgan fingerprints combined with Expert descriptors—consistently achieve the highest predictive accuracy and should serve as essential baselines for the development of new, more sophisticated models. In contrast, some current graph-based models, including AGILE, Chemprop, and KPGT, tend to show comparatively lower accuracy. The presented framework provides a standardized, transparent, and comprehensive benchmarking resource that enables meaningful comparison of emerging architectures and establishes strong baselines for future development of predictive models in lipid-based RNA delivery.
Related
- Influence of Lipid Thioether Modification and Formulation Method on Lipoplex Structure, Stability and Transfection Efficiency in HEK293T Cells
- A machine learning framework for solar cell design optimization and efficiency prediction
- Predicting nanoparticle tumor delivery efficiency using explainable machine learning for cancer nanomedicine design
- Benchmarking Faradaic Efficiency and O2 Quantification during Water Oxidation under Mild Conditions Using a 3D-Printed Cell and Clark Electrode
- Mechanistic Framework for Multicomponent Nanoparticle Assembly: Predicting RNA-lipid and PEI-DNA nanoparticle assembly
BibTeX
@article{mehradfar2026machine,
title = {{A machine learning benchmarking framework for lipid nanoparticle transfection efficiency prediction}},
author = {Asal Mehradfar and Mohammad Shahab Sepehri and José Miguel Hernández-Lobato and Glen S. Kwon and Mahdi Soltanolkotabi and Salman Avestimehr and Morteza Rasoulianboroujeni},
journal = {Communications AI \& Computing},
year = {2026},
month = jul,
doi = {10.1038/s44488-026-00007-x},
url = {https://doi.org/10.1038/s44488-026-00007-x},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - A machine learning benchmarking framework for lipid nanoparticle transfection efficiency prediction
AU - Asal Mehradfar
AU - Mohammad Shahab Sepehri
AU - José Miguel Hernández-Lobato
AU - Glen S. Kwon
AU - Mahdi Soltanolkotabi
AU - Salman Avestimehr
AU - Morteza Rasoulianboroujeni
PY - 2026
DA - 2026/07/14
JO - Communications AI & Computing
DO - 10.1038/s44488-026-00007-x
UR - https://doi.org/10.1038/s44488-026-00007-x
AB - The discovery of new ionizable lipids for efficient lipid nanoparticle (LNP)-mediated RNA delivery remains a major bottleneck in RNA therapeutics development. Recent advances demonstrate the potential of machine learning (ML) models to predict transfection efficiency directly from lipid structure, enabling high-throughput virtual screening and accelerating lead identification. However, as new models for LNP transfection prediction continue to emerge, the lack of rigorous and standardized benchmarking poses a significant risk and may undermine confidence in their reliability for discovery. Here, we present a robust ML benchmarking framework for evaluating transfection prediction models based on ionizable lipid structures. This framework systematically benchmarks diverse molecular representations—including Morgan fingerprints, Expert RDKit descriptors, and Grover graph-based embeddings—paired with a broad range of ML architectures spanning traditional models, feedforward neural networks, and state-of-the-art graph-based methods. In addition, the presented framework supports assessment of model generalization through Murcko scaffold splitting and evaluates prediction reliability beyond standard regression metrics by incorporating analyses of relative error distributions and ranking accuracy. Using a curated dataset of 1100 unique ionizable lipid structures derived from the HeLa transfection dataset originally reported by Xu et al. 1 , we show that within this framework, models leveraging explicit molecular substructure encoding—particularly multilayer perceptrons (MLPs) trained on Morgan fingerprints combined with Expert descriptors—consistently achieve the highest predictive accuracy and should serve as essential baselines for the development of new, more sophisticated models. In contrast, some current graph-based models, including AGILE, Chemprop, and KPGT, tend to show comparatively lower accuracy. The presented framework provides a standardized, transparent, and comprehensive benchmarking resource that enables meaningful comparison of emerging architectures and establishes strong baselines for future development of predictive models in lipid-based RNA delivery.
SN - 3091-292X
ER -