DeepMind Unveils AlphaGenome 2, Predicting Cell-Specific DNA Variant Effects

At a glance

Google DeepMind has released AlphaGenome 2, an AI model that predicts which cells a DNA variant affects through gene activity. It is an update to the first AlphaGenome and learns from single-cell data to reflect differences between cells.

Key points

1. 1

   AlphaGenome 2 predicts variant effects across hundreds of human and mouse cell types.

2. 2

   It learns from single-cell data, which measures gene activity in individual cells.

3. 3

   It reproduced a result that took a research team 1.5 years of experiments in about 30 minutes.

4. 4

   Non-commercial researchers can use it for free, while commercial use is offered as a private early-access program.

Why it matters

It could help researchers find disease-causing DNA variants faster, speeding up genetic disease research.

Also reported by

• The VergeGoogle’s AlphaGenome 2 zooms in on the cellular roots of disease
• Google DeepMind blogAlphaGenome 2: Deciphering how cells read DNA to better understand disease
• Unite.AIGoogle DeepMind의 AlphaGenome 2, 세포 상태별 변이 효과 예측
• 사이언스모니터구글 딥마인드 ‘알파게놈 2’ 공개…유형별 DNA 변이 효과 예측