"Multiscale Integration in Biological Systems Course at Institut Curie"

🔬 Flipping Life’s Chemistry: Exploring Mirror-Image Molecular Biology 🧬 What if we could build life as a perfect mirror image of what we know today? Mirror-image molecular biology is no longer science fiction — researchers are synthesizing mirror DNA, mirror enzymes, and even pieces of mirror ribosomes. In my latest blog, I break down: ✅ The science of chirality and stereoisomerism ✅ Key breakthroughs in mirror nucleic acids, enzymes & translation systems ✅ Potential applications in medicine, diagnostics, and sustainability ✅ The global debate on biosafety, ethics, and regulation This field doesn’t just challenge our technical skills — it forces us to ask deep questions about what life could be and how far we should go in redesigning it. 📖 Read the full article here: [https://lnkd.in/dw4_Umw5] 💬 I’d love to hear your thoughts: Do you think we should pursue the creation of self-sustaining mirror organisms, or put a moratorium until global governance is ready? #SyntheticBiology #MolecularBiology #Biotech #MirrorBiology #LifeSciences #ResearchEthics #ScienceCommunication

The ultimate playlist for AI in biology is here! 🎧🧪 Thrilled to see MIT's MLCB24 lecture series available online. Covering everything from protein language models and molecular generation to disease mechanisms and comparative genomics. A huge resource for learning how #MachineLearning is transforming life sciences. #ComputationalBiology #AI #GenAI #Bioinformatics #DrugDiscovery

A failed experiment. A persistent PhD student, Mikko Poutanen. A postdoc who took the lead, Matti Virkki. And eventually, a whole research line that grew around their work. That’s the story of OPTOSENSE: a journey that took us from fundamental studies on azobenzene photochemistry to humidity sensor prototypes used for monitoring the drying of concrete. From blue-sky research all the way to commercialization attempts. We didn’t establish a company, but we gained serendipitous discoveries, skills one cannot learn in a purely academic setting, and a new way of thinking about research and impact. I am deeply grateful to Mikko, Matti, Sami Vesamäki, Jani Patrakka, and many others who shaped this journey. OPTOSENSE may not have revolutionized humidity sensing, but it marked our first step towards supramolecular translation, broadening the way we approach science in our group. And that is success. For more stories like this, see our recent Editorial in Chemical Science Journal "Making molecules work – stories of supramolecular translation", with inspiring accounts by Anthony Davis, Jennifer Hiscock, Calden Carroll, Michael Haley, Darren Johnson, Jonathan Arambula, Krystle Karoscik, and Jon Sessler, ranging from diabetes management to antimicrobial materials. Many thanks to Stephen Goldup, Kate Jolliffe, Carri Cotton, and others at Royal Society of Chemistry for making this possible. And finally, thanks yet again to Ekaterina Osmekhina / microZOOMER for the brilliant comic that illustrates the OPTOSENSE journey... which still continues! https://lnkd.in/dk6Sxwmt

When shape meets function: stories from the structural biology frontier In biology, shape is everything. A protein fold determines how it interacts, signals, catalyzes, or misfires. From enzymes that accelerate life’s chemistry to membrane proteins that control the flow of information across cells, structure is the foundation of function. What excites me most about structural biology is not just the static snapshot of atoms in place, but the stories those structures tell about flexibility, disorder, dynamics, and evolution. With every new crystal, cryo-EM map, or computational model, we uncover how biology solves problems with elegance at the molecular scale. As structural biologists and biochemists, we are not just solving structures. We are piecing together the language of life, shapes, surfaces, and motions that hold the key to drug discovery, disease understanding, and the future of biotechnology. Whether it is a perfectly ordered fold or a dynamic disordered region, each structure has a story worth telling. Which molecular story fascinates you most? #StructuralBiology #Biochemistry #ProteinStructure #MolecularBiology #LifeSciences #DrugDiscovery #CryoEM #XrayCrystallography #ScienceCommunication #ResearchInnovation

Warning "static snapshot of atoms in place" still remains key! In reference to the interesting post below. I indeed agree that a dynamic view might be essential in some cases and we should always keep it in mind when working in structural biology. However we should never forget that enzymes rely on a highly precise atom positioning to allow both substrate binding and to create the electronic environment required for catalysis. Binding (substrate, effector, co-factor, other protein or drug) does also require fine positioning of "atoms in place". A snapshot of "atoms in place" may still go a long way! #structuralbiology #pdb #enzymestructure So, your thoughts ?

We’re excited to announce our 3rd year of partnership with “Frontiers in Systems Biology”, and cross-title collaborations with “Frontiers in Bioinformatics” and “Frontiers in Synthetic Biology” to provide an opportunity for you- past and present iGEM Competition teams- to publish your iGEM projects in peer-reviewed scientific journals! You can publish your research work on the following four topics: - Systems Biology and Synthetic Approaches in Achieving Sustainable Action; - Systems Biology in Biomedical Innovations and Healthcare; - Technological Advancements in Engineering: From Production to AI and Software; - Systems Biology in Healthcare and Pandemic Preparedness. For all topics, the Manuscript Summary Submission Deadline is November 30, 2025. For all topics, the Manuscript Submission Deadline is May 30, 2026. Learn more about the topics and the submission process on our website: https://lnkd.in/dC6jbS4N Frontiers-Bioinformatics and Physiology | Thomas C. Collin, PhD | Sean Manion | Yoram Vodovotz | Seán O'Donoghue

Plot twist: The most complex systems aren't in Silicon Valley. They're inside you. Your body runs ~37 trillion cells, each executing thousands of chemical reactions per second, all perfectly coordinated without a single line of code or project manager in sight. As an MSc Systems & Computational Biology student, I spend my days trying to decode this biological masterpiece. Here's what fascinates me: The scale is mind-boggling: A single cell contains more molecular interactions than there are web pages on the internet. Yet somehow, it all works in harmony. The resilience is incredible: Biological systems self-repair, adapt to stress, and optimize themselves in real-time. Netflix crashes if too many people log on—your liver processes toxins 24/7 for decades. The efficiency is unmatched: Your brain uses 20 watts of power (less than a light bulb) to outperform supercomputers in pattern recognition and decision-making. My goal? To understand these systems well enough to help when they go wrong. Whether it's cancer cells ignoring growth signals, neurons losing connections in Alzheimer's, or immune systems attacking the wrong targets—computational approaches are helping us see patterns that were invisible before. What biological "feature" amazes you most? #SystemsBiology #Biology #Innovation #Healthcare #GradSchool #ComputationalBiology

🚀 Excited to share a new paper on making protein language models more interpretable with Alp Tartıcı and Russ Altman! “Paying attention to attention: High attention sites as indicators of protein family and function in language models.” In this work, we highlight how attention weights can be turned into biologically meaningful insights: •⁠ ⁠Define High Attention sites from ESM-2 that spotlight residues tied to family & function •⁠ ⁠Show High Attention sites align with biologically important positions (e.g., active/functional sites) •⁠ ⁠Release High Attention annotations for the human proteome Now published in PLOS Computational Biology. 🔗 Open access: https://lnkd.in/gvn_ZBEb #Bioinformatics #ProteinLanguageModels #Interpretability #AIinBiology #Proteomics #ComputationalBiology #ESM2

RFdiffusion3: All-Atom Protein Design for Biomolecules The provided podcast is an excerpt from a scientific paper introducing RFdiffusion3 (RFD3), a novel deep learning diffusion model for the de novo design of protein structures and their interactions with other biomolecules. RFD3 is presented as an advancement over previous methods, shifting to an all-atom generative approach that explicitly models all polymer atoms, including side chains, to enable more complex and precise conditioning for design challenges. The paper details the architecture and training of RFD3, highlighting its efficiency and reduced computational cost compared to predecessors like AlphaFold3 and RFdiffusion2. Furthermore, it demonstrates RFD3's broad applicability through in silico benchmarks and experimental validation across tasks such as designing protein-protein, protein-nucleic acid, and protein-small molecule binders, as well as de novo enzyme design. The conclusion emphasizes that RFD3 offers unprecedented control and precision in designing complex biological functions due to its atomic-level foundation and comprehensive conditioning mechanisms. https://lnkd.in/gkafC__6

I am excited to share our recent work that has been published in ACS central science. In this paper, in collaboration with the labs of Gabriele Meloni and Hedieh Torabifard we explored the landscape of non-extant proteins to generate a novel funcional multidomain ATPase copper transporter using a combination of Latent Generative Landscapes, Molecular Dynamics and In vitro/ in vivo protein characterization. Briefly, we show that learning the sequence space of the P-type ATPase super family we are able to distinguish regions in a latent space that are specific to copper transport. Then using the generative properties of the landscape we are able to generate novel sequences that behave in our MD simulations similarly to wild type transporters. We then show that those multidomain proteins (>650aa) are able to fold, embed into membranes, have ATP hydrolysis and transport copper as originally designed., in some cases with higher hydrolysis rates. This project is the result of a great effort by three labs at the School of Natural Sciences and Mathematics, UT Dallas and showcases a methodology that is general and with potential applications to other proteins. One aspect to highlight is the complexity of the system, the multidomain nature of the protein and the fact that is a membrane protein with selective transport capabilities. Our proteins were >180 mutations away from the closest extant member of the family. This was possible due to the hard work of lead authors Fariha Hossain and Fernando Montalvillo Ortega. We were also happy that our article has made it to the cover where the 3D rendering of the landscape is made with sequence data generated by LGL :-) https://lnkd.in/e2iSmB8A