LMU researchers combine machine learning and molecular dynamics for new RNA delivery materials

by | Dec 1, 2025 | Health, Research

A team led by Professor Olivia Merkel from Ludwig-Maximilians-Universität (LMU) has developed an innovative platform that combines machine learning (ML) with molecular dynamics simulations (MD simulations). This method, called Bits2Bonds, enables the design and optimization of new polymerized materials for the targeted delivery of therapeutic RNA. The results were published in the Journal of the American Chemical Society.

The research was conducted as part of the ERC Consolidator Grant “RatInhalRNA”, which focuses on RNA formulations for lung application. Previous experimental screening of polymers has been time-consuming and expensive, while purely computational approaches fail due to a lack of data or high computational effort. Bits2Bonds overcomes these hurdles by combining coarse-grained MD simulations with ML-based molecule design. This allows thousands of potential carrier molecules to be virtually tested before they are experimentally tested, accelerating the development of safe and efficient RNA nanocarriers.

Credits: OpenClipart-Vectors/pixabay
Credits: OpenClipart-Vectors/pixabay

Olivia Merkel emphasizes that this hybrid strategy paves a new way to discover innovative materials for RNA therapeutics. The method supports a rational, high-throughput-based design and could bring personalized medicines closer. The team confirmed the simulation results by synthesizing and testing several polymer candidates for siRNA transport, where a clear agreement with biological efficacy was found.

The modular structure of the pipeline allows it to be adapted to other polymer types or nucleic acid therapeutics such as mRNA and CRISPR, opening up a wide range of applications. The work promises to significantly advance the development of RNA-based therapies, especially for pulmonary applications.

Original Paper:

From Bits to Bonds: High-Throughput Virtual Screening of Ribonucleic Acid Nanocarriers Using a Combinatorial Approach of Machine Learning and Molecular Dynamics | Journal of the American Chemical Society


Editor: X-Press Journalistenbüro GbR

Gender Notice. The personal designations used in this text always refer equally to female, male and diverse persons. Double/triple naming and gendered designations are used for better readability. ected.

X
Ich bin Invi, wie kann ich dir helfen?