🎉Big news! Our Nucleotide Transformer foundation models for genomics have been published in Nature Methods! NTs are foundational models for genomics, with up to 2.5 billion parameters, that are trained on genomes from 800+ species and 3000+ human individuals. Developed in collaboration with NVIDIA and the Technical University of Munich, NTs outperform specialized methods and other foundation models on a wide range of tasks in human genomics. NT models offer broad applications in genomics, from enhancing the prediction of splicing and regulatory mechanisms to improving variant effect prediction. 🚀 Highlights of our journey so far: ✅ 700,000+ downloads ✅ 120+ citations ✅ Now available open-source on Hugging Face for researchers and through DeepChain for enterprise users. NT is already making a significant impact in understanding the complexities of the human genome, and we’re just getting started. Here’s how NT is helping shape the next generation of genomic AI models: 👉 📘Paper: https://lnkd.in/ep9BTPuQ 📕 Research briefing: https://lnkd.in/ePxbZ2Kb 💻Tutorials (https://lnkd.in/eTyymuHu), models (https://lnkd.in/ehd3bDMd) and code (https://lnkd.in/e6P-NEEm) #AI #Genomics #NucleotideTransformers #NatureMethods #OpenSource #HuggingFace
Aman Yadav
2w
This credible work offers a more robust approach to leveraging genomic data...great work 👏
Awesome thanks for sharing
Head of Artificial Intelligence and Smart Mobility | Automotive, ML
2wI like what you do, and i would like to suggest the following: many of your followers like to have a short executive summary of your results without having to go through the whole paper or any other technical material. So instead of a picture of a green DNA sequence ( is that Hulk's DNA), i would like to see a comparative table of your results against say the other known player in this field, Deepmind. Keep the hard work.