Rezultati pretraživanja
  1. Proteins are essential to life. Predicting their 3D structure is a major unsolved challenge in biology and could impact disease understanding and drug discovery. I’m excited to announce that we have won the CASP13 protein folding competition!

    Prikaži ovu nit
  2. Although neural networks usually require massive datasets to do impressive things, for me the highlight of is the fact that it achieved state-of-the-art using only 30K training examples. Code: Paper:

  3. 3. pro 2018.

    Using AI to help scientists solve big questions is at the core of DeepMind’s mission. Today we’re delighted to announce our first significant milestone: a successful application of machine learning to the protein folding problem. (1/5)

    Prikaži ovu nit
  4. 15. sij

    Online : Improved protein structure prediction using potentials from deep learning

    Prikaži ovu nit
  5. Thanks to the CASP community for organising such a great benchmark, the gold standard for assessing protein folding techniques. Congratulations to the Science Team at on this fantastic achievement! ( is listed as ‘A7D’)

    Prikaži ovu nit
  6. 3. pro 2018.
    Prikaži ovu nit
  7. 17. sij

    Disappointed that didn’t get references quite right- first was

    Prikaži ovu nit
  8. 15. sij 2019.

    DeepMind's algorithm won a prominent protein folding competition last month, outperforming many well-known pharmaceutical companies. Its an amazing time to be alive! I'll explain how it works in this episode

  9. 2. pro 2018.

    enters protein folding: uses deep learning to predict 3D protein structure from sequence . Solving A structure is the beginning. Proteins are dynamic & function in different conformations. Curious to see details when paper is out.

  10. 12. pro 2018.

    can predict amazing protein structures via but we can refine to experimental accuracy via

  11. 5. pro 2018.

    So what's next, after ? We’re working to move beyond prediction to generate completely novel protein structures! Our lab's paper presents progress towards a deep learning-based complement/successor to for 1/n

    Prikaži ovu nit
  12. 17. sij

    because didn’t credit should I

  13. Understanding how proteins fold is a fundamental scientific question that could one day help unlock treatments for a range of diseases. Excited that our work, which uses deep neural networks to predict protein structures, is published in today!

  14. Majority won. Covered wide range of applications of Evolutionary Couplings and new approach to protein design. And team now also reading

  15. prije 18 sati

    has released a python package of with for protein fold prediction: contact prediction network, associated model weights and ~40Gb CASP13 dataset. Paper:

  16. 31. sij

    Top 👍 has released a package implementation of which include the contact prediction network, associated model weights and CASP13 dataset as published in Nature ( Authors access link : )👨‍💻

  17. 28. sij

    After the coffee break, we will continue at 3.30pm with Andrew Senior (, ), Katerina Vriza (, ) and last, but not least Katya Putintseva (@goneblotting, ). Looking forward!

    Prikaži ovu nit
  18. 26. sij

    Deep learning system "could model from scratch – i.e., based only on genetic sequence – better than any previous modelling system with a similar accuracy to systems drawing on templates of previously solved protein"

  19. Congrats to the team for publishing their exciting structure prediction work from CASP in and PROTEINS.

  20. 15. sij

    Two of our recent projects published in today! Our work on protein folding and another exploring the link between distributional RL and dopamine in the brain. Huge congrats to everyone who contributed to make these scientific achievements possible!

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.