Predicting Ternary Complexes for Targeted Protein Degradation

AlphaFold is no longer just predicting static structures; it is actively decoding the fundamental biophysics of protein dynamics. A compelling paper recently published in Physical Review Letters and available on bioRxiv, titled "Unifying constraints linking protein folding and native dynamics decoded from AlphaFold," explores the long standing question of how folding topology dictates native flexibility. By analyzing an extensive dataset of predicted structures, the authors uncovered a universal power law relationship demonstrating that long range contacts slowing down folding also act as physical anchors, systematically constraining conformational flexibility across different sizes and species. Understanding these topological and dynamic constraints unlocks incredible opportunities for computational biology and drug discovery. For instance, researchers can use contact order and predicted fluctuation entropy to select optimal starting conformations and define collective variables for enhanced sampling methods like metadynamics and alchemical free energy perturbation, drastically reducing computational time. Furthermore, this knowledge directly informs generative protein engineering pipelines, allowing specialists to precisely tune catalytic site flexibility versus scaffold stability without relying on rigid static models. It also provides a robust framework to map topological anchors against highly flexible regions, pinpointing allosteric hotspots and predicting the opening of cryptic cavities in dynamic targets. Beyond structural engineering, these insights profoundly impact virtual screening and evolutionary biology. Integrating topological constraint parameters directly into scoring frameworks and docking tools like Vina and Smina allows for a much more accurate evaluation of ligand binding across dynamic ensembles. Additionally, applying these scaling laws across whole proteome datasets enables the identification of aberrant flexibility profiles in mutated disease variants and the mapping of thermal adaptation strategies in extremophiles. Nature clearly trades rigid folding for dynamic function, and our computational tools are finally capturing this beautiful complexity. Read the full preprint and explore the methodology here: https://lnkd.in/dwytSp7Q

An AI pipeline that generates patentable genetic circuits. Here's what came back: 🧬 353 unique novel circuit architectures ✅ 100% passed ODE feasibility filtering 📋 29 circuits cleared prior-art checks — ready for provisional filing The system uses a Conditional Variational Autoencoder (CVAE) trained on a 24-promoter × 25-regulatory element × 24-gene vocabulary — 14,400 possible single-stage architectures. You give it a target biological behavior (bistability, toggle switch, AND gate, oscillator). It generates novel Promoter → Regulatory Element → Gene combinations that plausibly produce that behavior, scores each one with a Hill function ODE model, fingerprints them with SHA-256 for prior-art identity, and logs only the unseen, high-confidence ones as CANDIDATE_PATENT. Synthetic biology IP has historically required years of wet-lab iteration to produce a handful of novel constructs. Generative models collapse that to minutes. We're still in synthetic data territory — the next step is training on ODE-simulated circuits and crossing the 29 CLEAR candidates against iGEM. #SyntheticBiology #GenerativeAI #BiotechIP #CVAE #DeepLearning #GeneticCircuits #AIBiotech #Innovation

AlphaFold2 AI just cracked the protein folding code! Meet AlphaFold2, the game‑changer in biology. AlphaFold2 predicts 3D protein structures from amino‑acid sequences with near‑experimental accuracy, slashing months of lab work into minutes. It leverages deep learning and massive databases to model intricate folding patterns, unlocking insights into disease mechanisms and drug design. Researchers worldwide now can explore the proteome at unprecedented scale, accelerating everything from vaccine development to enzyme engineering. #AlphaFold2 #AIinBiology #ProteinFolding #DeepLearning #Biotech

AlphaFold2 AI just cracked the protein folding code! Meet AlphaFold2, the game‑changer in biology. AlphaFold2 predicts 3D protein structures from amino‑acid sequences with near‑experimental accuracy, slashing months of lab work into minutes. It leverages deep learning and massive databases to model intricate folding patterns, unlocking insights into disease mechanisms and drug design. Researchers worldwide now can explore the proteome at unprecedented scale, accelerating everything from vaccine development to enzyme engineering. #AlphaFold2 #AIinBiology #ProteinFolding #DeepLearning #Biotech

AlphaFold2 AI just cracked the protein folding code! Meet AlphaFold2, the game‑changer in biology. AlphaFold2 predicts 3D protein structures from amino‑acid sequences with near‑experimental accuracy, slashing months of lab work into minutes. It leverages deep learning and massive databases to model intricate folding patterns, unlocking insights into disease mechanisms and drug design. Researchers worldwide now can explore the proteome at unprecedented scale, accelerating everything from vaccine development to enzyme engineering. #AlphaFold2 #AIinBiology #ProteinFolding #DeepLearning #Biotech

In case you missed it... Check out this popular review published at the end of 2025 as part of our collection on Advances in Computational Protein Design, Structural Biology, and Drug Discovery! "Peptide-based drug design using generative AI" by Srinivasan Ekambaram and Nikolay Dokholyan Here, the authors review recent progress in peptide-based drug design using AI, focusing on generative architectures and interactions. They then examine AI-driven screening and delivery optimization for these peptide-based discoveries. Finally, they discuss the current limitations, practical challenges, and future direction with particular emphasis on data quality and autonomous drug discovery. Read the #OpenAccess Highlight article here 👉 https://lnkd.in/eNBprsKd

Finally, it was said out loud! This was a very insightful interview with pioneers of the virtual cell biology field. Biology is more complex than Perturb-seq readouts. The team generating more perturbation transcriptomics than anyone else I know is saying that is not enough to build causal frameworks from static cell state measurements alone. That is the most credible endorsement of the data gap I've seen all year. My experience building commercial strategies for omics data: Time isn't one more axis. It's the axis that makes the others interpretable. Metabolic flux reorganizes in minutes. Phospho-signaling in minutes. Proteome remodeling in hours to days. Sample once and you are integrating across four different clocks, then calling the average a cell state. Measuring proteins, metabolites and lipids -longitudinally- provides the functional context to learn pathway dynamics. #multiomics #proteomics #metabolomics #virtualcell #techbio #drugdiscovery / Latent Space | AI Engineering News, Essays and Interviews

What if we could simulate the entire biology of a cell, and predict how a potential drug or genetic change would affect it? This is exactly what a team of scientists is working on, with the goal of creating a "virtual cell" that can take multiple kinds of biological data into account and predict the effects of a series of interventions. Here are 3 key insights from their research: 1. The virtual cell could streamline biology and help researchers screen ideas earlier, focusing experiments on the most promising candidates and catching risks sooner. 2. The system would combine models for DNA, proteins, cells, tissues, and health outcomes into a dynamic simulator, allowing scientists to run experiments on biology before attempting them in the real world. 3. The near-term vision is not to replace animal studies or clinical trials, but to use AI as a filter: eliminate weaker drug ideas before researchers spend years and large sums testing them. So, can a virtual cell really revolutionize the field of medicine? #ArtificialIntelligence #HealthcareTechnology #InnovationInMedicine

I am pleased to share that I have co-authored my first peer-reviewed publication. Our mini-review, “Strengthening global biosecurity for synthetic nucleic acid technology: from sequence screening to risk-based governance in the AI era”, has been published in Frontiers in Bioengineering and Biotechnology. As synthetic nucleic acid technologies and artificial intelligence continue to advance, the limitations of sequence screening alone have become increasingly apparent. In this work, we examine why a broader, systems-based and risk-informed framework is needed to support responsible innovation while reinforcing global biosecurity. I am deeply grateful to the SynBioForAll team for the opportunity to contribute to this important discussion. Their guidance, collaborative spirit, and commitment to advancing the field have made this milestone possible, and I am proud to have been part of the process. You can read the full article here: https://lnkd.in/eEGzwurP Additional research contributions from the team are available at: https://lnkd.in/eK-f6qnh I welcome any thoughts or reflections from colleagues working at the intersection of synthetic biology, AI, and biosecurity. #SyntheticBiology #Biosecurity #ResponsibleInnovation #AIinBiology #SynBioForAll

Predicting multi-entity biomolecular complexes remains a computational bottleneck in structural biology. Vecura now hosts #Protenix, an open-source AlphaFold 3 reproduction. It predicts full 3D atomic structures of protein, DNA, RNA, and ligand complexes using a diffusion-based decoder and Pairformer trunk, returning mmCIF files with rigorous confidence metrics like pLDDT and ipTM. This enables scalable protein modeling and molecular discovery, allowing researchers to evaluate complex binding modes with inference-time scaling. Vecura provides a secure, no-code Agentic AI platform to test and deploy Protenix without managing complex GPU infrastructure or MSA preprocessing. Read more in the Vecura blog: https://lnkd.in/gupQ8Sm8 Many thanks to: Protenix Team - Yuxuan Zhang, Chengyue Gong, Hanyu Zhang, Wenzhi Ma, Zhenyu Liu, Xinshi Chen, Jiaqi Guan, Lan Wang, Wenzhi Xiao, authors of Protenix. #Vecura #AIForScience #ProteinModeling #MolecularDiscovery #ComputationalBiology