Biohub Cell Tracking Competition on Kaggle

Biohub & Kaggle offer $60K for #DataScience teams to build machine learning models that detect, track, and link cells across time using real 3D microscopy data, including identifying cell divisions and reconstructing cell lineages. https://lnkd.in/dTNad6TD

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

23.5% of the library, most of the map These authors made 203 formulations in the lab and predicted the other 661. They paired a parallel microfluidic chip that makes many tiny LNP batches at once with a machine learning model, then let biology pick the winner. What did they find? 1) Size prediction was accurate enough to rank a virtual library: MAE under 7 nm, correlation 0.94. 2) One rule held across all four ionizable lipids: more PEG-lipid, smaller particles, but less uniform. CKK-E12 broke the pattern on concentration, which is the sort of thing a broad screen exists to find. 3) It delivered. Every loaded formulation came in under 100 nm, CKK-E12 reached ~92% gene knockdown, and it also showed the strongest lung signal in mice. Code is on GitHub. Some issues? For sure. The uniformity (PDI) model was weak, and the authors say so and show the diagnostics rather than burying it. Characterization was triplicate from one prep, no separate batches. The mouse work tracks a dye at 2h, so it shows where particles went, not what they did. And the best performer was the one with the worst physical properties, which undercuts the premise a little. The lesson to keep in mind? Size and uniformity are gates, not goals. The pipeline still works; you just need the biology at the end. Read more: https://lnkd.in/gmuqxe8x #LNP #siRNA #DrugDelivery #MachineLearning #FormulationScience #Nanomedicine

AlphaFold 3 AI just cracked protein folding—again! Meet AlphaFold 3, the next leap in bio‑tech. AlphaFold 3 expands beyond static structures, predicting protein dynamics and interactions in real‑time. It integrates multimodal data, from cryo‑EM maps to cellular environments, delivering atom‑level accuracy for complexes previously deemed unsolvable. This speed and precision slash experimental costs, accelerating drug design and enzyme engineering. By democratizing high‑fidelity models, it empowers labs worldwide to explore biology at an unprecedented scale. #AlphaFold3 #AIinBiology #ProteinScience #BioTech #FutureOfMedicine

AlphaFold 3 AI just cracked protein folding—again! Meet AlphaFold 3, the next leap in bio‑tech. AlphaFold 3 expands beyond static structures, predicting protein dynamics and interactions in real‑time. It integrates multimodal data, from cryo‑EM maps to cellular environments, delivering atom‑level accuracy for complexes previously deemed unsolvable. This speed and precision slash experimental costs, accelerating drug design and enzyme engineering. By democratizing high‑fidelity models, it empowers labs worldwide to explore biology at an unprecedented scale. #AlphaFold3 #AIinBiology #ProteinScience #BioTech #FutureOfMedicine

AlphaFold 3 AI just cracked protein folding—again! Meet AlphaFold 3, the next leap in bio‑tech. AlphaFold 3 expands beyond static structures, predicting protein dynamics and interactions in real‑time. It integrates multimodal data, from cryo‑EM maps to cellular environments, delivering atom‑level accuracy for complexes previously deemed unsolvable. This speed and precision slash experimental costs, accelerating drug design and enzyme engineering. By democratizing high‑fidelity models, it empowers labs worldwide to explore biology at an unprecedented scale. #AlphaFold3 #AIinBiology #ProteinScience #BioTech #FutureOfMedicine

The Future Arrived in Barcelona ⭐ : Please welcome the #Pyxa from Stellaromics Inc. Spatial Genomics in 3D ✨ Delivering on our roadmap to apply highest-resolution 2D/3D Spatial Genomics technologies at the Centro Nacional de Análisis Genómico (CNAG). "Spatial Genomics is moving at an incredible pace! We have just seen product launches for full transcriptome 2D profiling on the Atera and CosMx, but this seems yet again to be only the warm-up phase. Multiple companies are approaching now a resolution in 3D. At this speed of launches and given the competitive landscape, we will appreciate the full potential of technology innovation in the very near future." Want to learn more about Spatial Genomics instruments and Digital Pathology application? 👉 Check this out: https://lnkd.in/ejBMrjGa Human Cell Atlas, Omniscope, Parc Científic de Barcelona

3D Is The New 2D ⭐ Three-dimensional #Spatial Genomics made commercially available with the #Pyxa from Stellaromics Inc. Check out the Blog and Newsletter at https://lnkd.in/dCgGAMVt regular updates 📰

Summer reflections from the home office on #SpatialTranscriptomics. Our runner ducks 🦆 are happily purging the garden of slugs. The freezer is stocked with ice cream in anticipation of a second heatwave this weekend. Spring was busy at Transcriptomic Insights and - happily - the summer is too. Science can be a slow process. But the last five months have left me with a feeling that biological science is switching gears. That feeling has strong roots in the spatial transcriptomics field. Spatial transcriptomics keeps pushing limits in terms of resolution, throughput and sensitivity. Just look at two platforms launched this year: Atera from 10x and StrataMap from Illumina. Two different chemistries, both geared for whole transcriptome, high resolution and high throughput. Both with large capture areas. What gets me excited about advancements in the spatial field is a simple thought: Say we're after the transcriptome of individual cells in a tissue. What if we skipped the whole cell/nuclei isolation and microfluidics part, with all the biases it introduces and optimization it requires. And instead, we just slap the tissue on a slide - with the added benefit of spatial localization. That's becoming reality. And for many applications I believe spatial will soon become the go-to method for single-cell resolved insights in tissue. Of course, single cell/nuclei will always have their applications, such as profiling cell suspensions or gathering information across huge numbers of cells. But the value of measuring cells in their native tissue context is immense. Add to that the growing number of modalities that can now be layered onto spatial data. The acquisition of Proteintech Genomics by 10x is a clear indicator of where the field is heading. Imagine a future where every cell comes with information about location, gene and protein expression and more (epigenomics, bacterial and viral infiltration, post-translational modification, chromatin conformation – it's all being worked on!) All that, from a slice of tissue. What a time to be in science!

What if your sequencer could decide what to read while it reads? ⚡ Satrio Wibowo (LongTREC) is up next to improve adaptive sampling for transcriptomics, steering nanopore in real time to capture more of what matters. #VALT2026 #LongReads #AdaptiveSampling #Nanopore