1PB dataset precomputes all ~9 billion possible SNVs across the entire human genome
anshulkundaje · x · 2026-09-08
A retweeted post announces a massive genomics resource: variant effect predictions precomputed for all 9 billion possible single-nucleotide variants (SNVs) across 100% of the human genome, totaling 1 petabyte.
For comparison, AlphaFold DB (AFDB) is only 23-30 TB and covers the 1.5-2% protein-coding portion of the genome via 200M protein structures — making the new dataset over 30x larger in scale and far broader in coverage. Retweeter Anshul Kundaje (Stanford genomics professor) calls it huge.
More from Research
- DeepMind launches AlphaGenome Atlas: precomputed impact predictions for ~9B genome variants — anshulkundaje · 2026-09-09
- Paper: general coding agents beat purpose-built data agents by up to 37 points — RishiBommasani · 2026-09-09
- OpenAI claims 10,000 concurrent agents produced a proof resolving the Navier–Stokes Millennium Prize problem — scaling01 · 2026-09-09
- OpenAI claims 10,000 agents solved Navier–Stokes Millennium Prize problem in 88 hours — scaling01 · 2026-09-09
- GPT-6 Astra cracked 10 'open' Erdős problems and improved a 1970s bound — stuffyokodraws · 2026-09-09
- New evidence backs the hypothesis that AI tools benefit skilled users most — srchvrs · 2026-09-09