Researcher

Augustin Žídek

Augustin Žídek is a Staff Research Engineer at DeepMind who works on AlphaFold, an AI system for protein-structure prediction. He studied mathematics and informatics at Cambridge, earned bachelor's and master's degrees in 2016, and joined DeepMind in autumn 2016. A 2018 paper on reinforcement learning preceded his work on AlphaFold structure prediction, the public AlphaFold Database, and AlphaFold 3's predictions of molecular interactions. In 2025, he coauthored a paper on new AlphaFold Database and 3D-Beacons capabilities. In 2026, he contributed to DeepMind's SynthID Bio research article.

Works
27
Citations
80,544
h-index
17

Publication numbers from OpenAlex, as of 18 October 2026.

Timeline

  1. 2012–2016

    Computer science at Cambridge

    He attends Gymnázium Frýdlant nad Ostravicí, then studies mathematics and informatics at Cambridge, completing bachelor's and master's degrees in 2016.

  2. 2016–2017

    Joining DeepMind’s folding team

    He joins DeepMind in autumn 2016 and is part of its protein-folding team from 2017.

  3. 2018

    Transfer learning for reinforcement agents

    In 2018, he coauthors “Transfer in Deep Reinforcement Learning Using Successor Features and Generalised Policy Improvement,” a paper on reusing skills across tasks.

  4. January 2020

    AlphaFold’s early structure model

    In January 2020, he is named among DeepMind’s collaborators on AlphaFold, its system for predicting a protein’s 3D structure from genetic sequence.

  5. July 2021

    Protein structures at proteome scale

    In July 2021, he coauthors a DeepMind report on applying AlphaFold at proteome scale and sharing predictions through the AlphaFold Database.

  6. May 2024

    AlphaFold 3 predicts interactions

    In May 2024, he coauthors the AlphaFold 3 paper, which describes joint-structure predictions for complexes containing proteins, nucleic acids, small molecules, ions and modified residues.

  7. 2025

    New data for AlphaFold DB

    In 2025, he coauthors “AlphaFold Protein Structure Database and 3D-Beacons: New Data and Capabilities.”

  8. As of October 2026

    Staff Research Engineer at DeepMind

    In September 2026, he contributes to Google DeepMind’s SynthID Bio research article on watermarking for synthetic biology. As of October 2026, he is a Staff Research Engineer at DeepMind and works on AlphaFold.

Career

  1. Since 2016Google DeepMindStaff Research Engineer

Education

  1. 2016University of CambridgeBachelor's and master's degrees in mathematics and informatics

Most-cited works

  1. Highly accurate protein structure prediction with AlphaFold2021 · Nature · 43,412 citationsWith Alex Bridgland, Alexander Pritzel, Andrew W. Senior, Anna Potapenko, Bernardino Romera-Paredes and Clemens Meyer
  2. Accurate structure prediction of biomolecular interactions with AlphaFold 32024 · Nature · 15,516 citationsWith Alex Bridgland, Alexander Pritzel, Anna Potapenko, Demis Hassabis, John Jumper and Josh Abramson
  3. AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models2021 · Nucleic Acids Research · 7,786 citationsWith Demis Hassabis, John Jumper, Michael Figurnov, Pushmeet Kohli and Stig Petersen
  4. Protein complex prediction with AlphaFold-Multimer2021 · 3,780 citationsWith Alex Bridgland, Alexander Pritzel, Andrew W. Senior, Anna Potapenko, Demis Hassabis and John Jumper
  5. Improved protein structure prediction using potentials from deep learning2020 · Nature · 3,372 citationsWith Alex Bridgland, Andrew W. Senior, David Silver, Demis Hassabis, John Jumper and Karen Simonyan

Sources and credits

Photo: We have no published photo of Augustin Žídek yet. Send us a free-licensed photo.

  1. Augustin Žídek | Nadace The Kellner Family Foundation kellnerfoundation.cz. Checked 18 October 2026.
  2. Vývojář: Vymýšlíme umělou inteligenci tak, aby vyřešila všechno ostatní seznamzpravy.cz. Checked 18 October 2026.
  3. Trainer biographies | AlphaFold Education Summit ebi.ac.uk. Checked 18 October 2026.
  4. Transfer in Deep Reinforcement Learning Using Successor Features and Generalised Policy Improvement proceedings.mlr.press. Checked 18 October 2026.
  5. AlphaFold: Using AI for scientific discovery deepmind.google. Checked 18 October 2026.
  6. Enabling high-accuracy protein structure prediction at the proteome scale deepmind.google. Checked 18 October 2026.
  7. Accurate structure prediction of biomolecular interactions with AlphaFold 3 nature.com. Checked 18 October 2026.
  8. Publications – Augustin Žídek augustin.zidek.eu. Checked 18 October 2026.
  9. SynthID Bio: Watermarking methods for synthetic biology — Google DeepMind deepmind.google. Checked 18 October 2026.
  10. Augustin Žídek augustin.zidek.eu. Checked 18 October 2026.