LNP Space

LNPSpace: An AI-Assisted Literature Platform for Lipid Nanoparticle Research

Efficient delivery remains one of the major challenges in the development of nucleic-acid-based therapies, including mRNA therapeutics, gene therapy, and genome editing. Lipid nanoparticles (LNPs) have emerged as one of the most important delivery platforms in this field, but the rapid expansion of LNP research has also created a fragmented and increasingly difficult-to-navigate literature landscape.

LNPSpace was developed as a pilot-scale, AI-assisted literature platform designed to support the retrieval and exploration of scientific information related to LNP formulations and nucleic acid delivery.

The platform combines scientific literature retrieval with a Retrieval-Augmented Generation (RAG) workflow. Relevant open-access publications are identified, processed at the section level, and searched for evidence directly related to a user's question. Retrieved information is then used to generate concise, literature-grounded responses linked back to the supporting publications.

A central feature of LNPSpace is the structured retrieval of LNP formulation information. Users can explore reported lipid compositions, nucleic acid cargoes, particle characteristics, formulation parameters, and experimental applications where these data are available. Information from multiple studies can also be organized into comparative formulation tables, helping researchers evaluate reported LNP systems without manually extracting data from individual publications.

LNPSpace additionally includes a Talk-to-PDF function. Once a relevant publication has been identified, users can select that article and perform publication-specific queries to examine experimental methods, formulation details, results, or other information directly within the paper.

For the GenE-Humdi community, such literature tools may be particularly useful at the interface between genome editing and delivery technology. As increasingly sophisticated genome- and epigenome-editing systems are developed, identifying appropriate delivery strategies remains essential for translating these technologies toward therapeutic applications. AI-assisted literature retrieval can help researchers more rapidly connect emerging editing approaches with relevant delivery systems and formulation knowledge reported across the field.

LNPSpace is currently a pilot-scale platform with coverage focused on a limited recent literature window. Future development will expand the searchable literature corpus, improve automated extraction and standardization of formulation-level information, strengthen benchmarking and evidence evaluation, and extend the platform toward additional areas of nucleic acid delivery and nanomedicine.

The project was developed with support from an e-COST Virtual Mobility Grant within GenE-Humdi, providing an opportunity to explore how AI-assisted knowledge retrieval can complement conventional literature searching and support research at the intersection of genome editing, nucleic acid delivery, and computational technologies.

Explore LNPSpace: LNPSpace — Literature Review Platform