The work has occupied a large fraction of my lab for past two years, included both huge modeling effort (by students/postdocs J Evans, M Sample S Roy, M Matthies) and experiments (H Liu and T Diep) We also worked iwth J Hihath and Y Ke's lab for help with carbon nanotube scaffolding
Finally, the cubes can have multiple applications in nanoscale manufacturing. For example, they can be used to create 3D carbon nanotube circuits, as shown below where a cube is used to scaffold a 3D carbon nanotube junction for scaling up nanoelectronics manufacturing to 3D:
and it can also be used to design dynamic reconfigurable structures, using "cube displacement" reaction, a 3D analogue of the often-used DNA strand displacement used for molecular computing:
The platform can also be used to make crystals, such us the Sponge crystal:
Menger cube in oxDNA (left) and experimental characterization with SEM (scale bar 100 nm)
Our modeling platform finds "minimal assembly kit", the smallest number of cubic shaped DNA origami blocks, interacting with single-stranded overhangs, that are needed to reliably assemble into target prescribed 3D shape. To demonstrate of our platform we build Menger cube (1.9 GDa) fractal:
🧪🧬 You can now play Minecraft with DNA nanotechnology! In our latest preprint we introduce DNA polycubes platform: A modeling driven system, where we use SAT-solvers+multiscale simulations to design 3D finite-size structures that assemble in high yield
chemrxiv.org/doi/full/10....
As has become our yearly tradition with John Russo and his team at Sapienza University, there is always a Christmas-themed video based on current patchy self-assembly problem that we work on. The problems (and the videos!) get more sophisticated every year: www.youtube.com/watch?v=keZI...
We are happy to announce the oxDNA workshop 2026 at Arizona State University in February 16th - 20th . It will be a three day conference followed by a two-day hackathon of coding up new features in oxDNA and associated tools, as well as tutorials on its usage:
sulcgroup.github.io/oxdnaworkshop/
Thank you @mskcancercenter.bsky.social for highlighting our work on #viralmimicry in our genome, its evolution, the selective forces on it, and its potential role in cancer.
A great years long collaborative effort with @petrsulc.bsky.social, @daniel-decarvalho.bsky.social, & many great colleagues!
While this work presents a first step towards 'decoding the repeatome' in human genome, it also presents a fast an accessible framework to assess stimulatory potential of transcripts coming from various experimental datasets.
Complementary to this hypothesis, it is also possible that maintaining these patterns is also beneficial for the repeat families, as we argue in the case of the highly studied family of LINE-1 repeats, where high CpG content might help prevent their deleterious effect on organism's fitness.
We argue that these repeats might have been co-opted to trigger innate immune system response upon translational dysregulation, such as in cancer cells. We checked experimentally that repeat transcripts that we marked as anomalous do bind to pattern recognition receptors of innate immune system.
It allows us to quantify how anomalous given sequence is with respect to what one would expect from a typical transcript from human genomes. The two chosen motifs, CpG content and double-stranded regions,are hallmarks of RNA viruses and our immune system recognizes them as immunostimulatory targets.
In our work, we instead focus on developing statistical mechanics-based framework to identify anomalous sequence (CpGf) and structure (long double-stranded segment) motifs in transcripts of these regions, which are known to be recognized by innate immune system receptors.
A majority of our genome does not code for proteins nor has any established regulatory function. These regions are sometimes called 'junk DNA'. Their function, or lack of, is a very active topic of research.
🧪 In our new work (just published in Cell Genomics) with Benjamin Greenbaum , John LaCava, Daniel De Carvalho, PhD , Simona Cocco and Remi Monasson labs, we venture into the dark matter of human genome:
www.cell.com/cell-genomic...
🧵A shot thread summary below in comments:⬇️
On this first day of #DNA31, reminder for the mol pro community that I put together this feed which picks up the last 3 days of posts from our community! Enjoy the meeting! 🧬
A great effort to chart possible future paths for the molecular programming field
The predictions were verified experimentally. Our results have implications for DNA/RNA molecular computing, design of DNA computing systems that interface with RNA triggers, and as we show in our presented model, it can also have implications for predicting the kinetics in CRISPR-based systems.
It shows that just by permuting the distribution of bases in the duplex (while keeping AT / CG base pair number constant), the kinetics of the reaction can be altered by orders of magnitude, stemming from the details of differences in stability between hybrid and canonical bases in DNA/RNA systems.
It is also of importance for biological systems like CRISPR-Cas9. In this collaboration, Eryk Ratajczyk supervised by Louis, Doye and Turberfield groups from Oxford, uses our new oxDNA-oxRNA hybrid coarse-grained model to study the strand displacement RNA invades DNA duplex.
🧪🧬
Our new paper is just out in PNAS:
www.pnas.org/doi/10.1073/...
It is about nucleic acid strand displacement, which is a key reaction in molecular computation, and part of the success of the oxDNA model comes from the fact it can efficiently simulate this process.
1/4🧵
Great work by our student Navraj and Subhajit, building on top of the ground work of previous database by @floppleton.bsky.social and Michael Matthies
Just in time for #FNANO conference in Snowbird, UT, our updated and improved version of nanobase.org server is online. Just like the PDB database helped to share protein structures, we want to create a community resource for the DNA/RNA/protein nanotechnology design researchers to share their work
4/4: Our crystal example thus provides a falsifiability test for classical nucleation theory, showing need for also taking into account structural fluctuations.
Excellent work by Camilla Beneduce and Diogo Pinto from John Russo's group in Rome!
3/4 However, we find that one of the three possible crystal phases is strongly preferred, due to the crystal lattice ordering being similar to the local ordering in the liquid phase.
2/4: The assumptions of CNT rely on the free energy difference between bulk phase and the melt, and the interfacial free energy. In this work, we look at a two component system that forms a polymorph: it can nucleate into three different lattices with same interfacial and bulk free-energy.
🧪In our new paper (journals.aps.org/prl/abstract...) in collaboration with Russo, Romano, Rovigatti and Sciortino groups in Rome / Venice, we look at Classical Nucleation Theory: a popular model of nucleation process, n is a key phenomena in self-assembly, self-organization and phase transitions.
Thank you Ulrich, it was a pleasure visiting Heidelberg and meeting you. Many thanks to the amazing group of @kgoepfrich.bsky.social for hosting me!
Great collaboration with @laklab-tubs.bsky.social , simulations carried out by my student Josh Evans