Sundar Pichai Says ‘Every Cell In Our Body Shares the Same DNA Blueprint’ as Google DeepMind Unveils AlphaGenome 2 Amid Drugmakers' AI Skepticism

Alphabet Inc.’s (NASDAQ:GOOG) (NASDAQ:GOOGL) subsidiary Google DeepMind announced AlphaGenome 2, an artificial intelligence (AI) model designed to predict how DNA variations affect specific cell types and help researchers investigate disease mechanisms.

AlphaGenome 2 was trained on more than 107 million individual cell profiles from human and mouse samples, the company said in a press release on Wednesday. Google DeepMind said the model identifies nearly twice as many genetic variants that change gene expression in cells as its previous state-of-the-art model.

AlphaGenome 2 Focuses On Specific Cell Types

The original AlphaGenome model made predictions across broad tissues and cell lines. AlphaGenome 2 takes a more targeted approach by allowing researchers to select a particular tissue, cell type, or biological state when studying a DNA sequence. This helps the model predict how the same genetic variation may affect different types of cells.

Professor Sarah Teichmann, co-founder of the Human Cell Atlas, said, "What makes AlphaGenome 2 exciting is how it adapts directly to new single-cell data without retraining, predicting how DNA variants act in newly profiled cell states from even a small sample of cells."

In a benchmark called TenK10K, which maps 28 primary immune cell types, AlphaGenome 2 identified 30% of causal genetic variants, compared with 16% for the original AlphaGenome model fine-tuned on single-cell data and other specialized tools. It also improved its ability to predict whether a genetic variant increases or decreases gene activity, with results of 43% compared with 23%.

Pichai Explains How Cells Read DNA Differently

In an X post, Google CEO Sundar Pichai explained the model’s purpose, writing, "Every cell in our body shares the same DNA blueprint, but different cell types (like heart muscle cells vs. immune cells) read that blueprint differently." He added that AlphaGenome 2 "predicts how these specialized cells interpret DNA, turning genetic code into biological insights."

AlphaGenome 2 is helping uncover the cellular drivers of disease, giving scientists a powerful tool to predict how DNA variants impact specific cell types.

Every cell in our body shares the same DNA blueprint, but different cell types (like heart muscle cells vs. immune cells)… pic.twitter.com/1ujnGcmcEO

— Sundar Pichai (@sundarpichai) November 11, 2026

Google DeepMind also demonstrated the model’s capabilities using an existing research finding involving inflammatory bowel disease (IBD), a condition that causes chronic inflammation in the digestive tract. A previous study linked a genetic variation to increased activity of the ETS2 gene in gut macrophages, which are immune cells found in the digestive system. Google DeepMind said AlphaGenome 2 reproduced the finding computationally, predicting that the variation specifically increases ETS2 activity in those cells.

The model is available for non-commercial research through the AlphaGenome API and in private preview for commercial research. Google DeepMind said it has not been validated or approved for clinical or diagnostic use.

AI’s Promise And Limits In Drug Discovery

Drugmakers have cautioned that AI cannot resolve biology’s unknowns on its own. AstraZeneca CEO Pascal Soriot described disease and biology as complicated, while Vertex executive chairman Jeff Leiden said roughly 50 AI companies tested against Vertex’s 10,000-compound dataset had not produced a working drug.

Last month, Meta Platforms Inc. (NASDAQ: META) and Alphabet units Google DeepMind and Isomorphic Labs said they committed $300 million to a broader effort to develop AI models that predict how cells respond to drugs and genetic changes. That initiative is separate from AlphaGenome 2.

Alphabet’s shares (GOOG) closed Wednesday’s regular trading session 2% higher at $290.63. The stock rose 0.26% to $291.39 in after-hours trading and is down 7.38% year to date.

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Image via Shutterstock