™AlphaGenome 2: Teaching AI to Read the Entire Human Genome — DeepMind Interview (39:46)
Official Google DeepMind interview, 39:46, with Hannah Fry speaking to Žiga Avsec and Adam Gayoso about AlphaGenome, the AlphaGenome Atlas and AlphaGenome 2, and how it reads non-coding DNA down to single-cell resolution to pinpoint disease-causing mutations.
Key points Tap a time to jump
Introduction to AlphaGenome: 98% of human DNA does not build proteins; it controls how cells work
Genotype, phenotype and DNA: how genes relate to traits
From Transformers to AlphaGenome: how the model architecture evolved
Mapping the genome
Rare disease cases: using AlphaGenome to analyse disease-causing mutations
AlphaGenome 2: reading non-coding DNA down to single-cell resolution to pinpoint disease-causing mutations
Future vision
Hannah Fry's closing thoughts
Comments
More to watch
All →- 0:57Introducing Claude Whiteboards, Sheets, and Data (0:57)
- 1:25:52Designing for Agents — Zeno Rocha, Resend (1:25:52)
- 2:43:49What Frontier Labs Buy: A Data Deep Dive with Handshake AI — Garrett Lord & Team
- 2:00:44[One Million AI Engineer Special] How Stanford Teaches AI-Native Engineering — Mihail Eric, Monaco
Related news
Today
Nov 12 Thursday- Brief
- GitHubGitHub Weekly Vol.5: 6 Projects Worth a Look This Week
Tested or read — weeklyAll →
- Quiz
- Call