As someone who spent their entire academic career, from B.Sc. to M.Sc. to Ph.D., within a Kavli Institute for Nanoscience Discovery (first in Delft and now in Oxford), I’ve had the privilege of seeing firsthand just how beautifully intricate the nanoscale world can be. Now, as my research focuses on lipid nanoparticles for genetic therapeutics and vaccines, I would like to use this platform to advocate for what I believe is one of the most transformative frontiers in modern medicine: the rational design of nanomaterials for therapeutic delivery.
Last month, I attended a conference in Cambridge aptly titled Integrating Chemistry and Engineering Biology to Create the Next Generation of Vaccines. One of the highlights was a keynote lecture by Prof. Neil King from the University of Washington, whose group is pioneering the de novo design of protein nanoparticles using state-of-the-art computational tools and machine learning approaches.
Traditionally, protein nanomaterials have largely been assembled by repurposing naturally occurring protein structures. What makes this new generation of research so exciting is that scientists are increasingly moving beyond repurposing biological structures toward designing entirely new nanoscale architectures from first principles. In their recent preprint, De novo design of protein nanoparticles with integrated functional motifs1, King and colleagues describe a computational pipeline that combines RFdiffusion2, ProteinMPNN3 and AlphaFold4 to generate self-assembling protein nanoparticles with atom-level precision (Fig. 1).
Figure 1. Computational pipeline for the de novo design of self-assembling protein nanoparticles. Machine learning-based tools including RFdiffusion, ProteinMPNN and AlphaFold were combined to generate and validate highly symmetric protein nanostructures with atom-level precision. Adapted from Haas et al., bioRxiv (2026).
The scale of this achievement is remarkable. The team generated tens of thousands of candidate protein building blocks in silico, assembled these into highly symmetric nanostructures, and experimentally validated many of them using cryo-electron microscopy and X-ray crystallography. Several of these nanoparticles matched their computational design models with deviations of only ~1 Å RMSD, approaching atomic accuracy.
What makes these systems particularly exciting for medicine is that they are not simply structural curiosities. These nanoparticles can be engineered to display functional biological motifs, carry therapeutic cargo or present viral antigens in geometries specifically optimized to stimulate immune responses (Fig. 2). In the preprint, the authors demonstrate the design of tailored nanoparticle vaccine scaffolds capable of eliciting robust antibody responses in mice.
Figure 2. Examples of antigen-tailored protein nanoparticles designed for vaccine applications. These engineered nanostructures can be customized to display viral antigens in geometries optimized for immune stimulation and were shown experimentally to elicit robust antibody responses in mice. Adapted from Haas et al., bioRxiv (2026).
This represents a fundamental shift in how we think about therapeutic design. We are beginning to engineer nanomedicines in much the same way civil engineers design bridges or architects design buildings: through iterative modeling, simulation, optimization and structural validation5,6—except now everything happens at the scale of a billionth of a meter.
For decades, nanomedicine promised revolutionary advances in targeted drug delivery, vaccines and molecular diagnostics. Yet progress was often limited by our inability to precisely control nanoscale structure, assembly and function7. The convergence of machine learning, structural biology, protein engineering and high-resolution imaging is finally changing that. We are now approaching an era where bespoke nanoparticles can be designed in silico for highly specific biological tasks, whether delivering mRNA, targeting tumors or presenting viral antigens with near-atomic precision.
As Prof. King put it during his talk: “We have finally reached the golden age of nanomedicine.”
And perhaps for the first time, that statement no longer feels aspirational. It feels technically achievable.
References
1. Haas, C. M. et al. De novo design of protein nanoparticles with integrated functional motifs. bioRxiv [Preprint] (2026).
2. Watson, J. L. et al. De novo design of protein structure and function with RFdiffusion. Nature 620, 1089–1100 (2023).
3. Dauparas, J. et al. Robust Deep Learning-Based Protein Sequence Design Using ProteinMPNN. Science 378, 49–56 (2022).
4. Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
5. Hou, X., Zaks, T., Langer, R. & Dong, Y. Lipid nanoparticles for mRNA delivery. Nature Reviews Materials 6, 1078–1094 (2021).
6. Mitchell, M. J. et al. Engineering precision nanoparticles for drug delivery. Nature Reviews Drug Discovery 20, 101–124 (2021).
7. Peer, D. et al. Nanocarriers as an Emerging Platform for Cancer Therapy. Nature Nanotechnology 2, 751-760 (2007).
