DeepMind's Co-Scientist tool helps researchers identify genetic factors that reverse cellular aging in human cells
A collaboration between Google DeepMind and biologists Omar Abudayyeh and Jonathan Gootenberg shows that AI can accelerate the discovery of rejuvenation pathways by analyzing genetic screens and scientific literature at scale.
1 source · cross-referenced
- DeepMind's Co-Scientist AI tool analyzed tens of thousands of scientific papers to identify over 20 novel genetic factors that could reverse cellular aging.
- Lab experiments validated at least two of Co-Scientist's proposed factors, successfully driving cells toward a younger, more functional state.
- Co-Scientist reduced the time needed to analyze large genetic screening data from up to six months to a few days by cross-referencing results with scientific literature.
- The tool addresses two key bottlenecks in aging research: identifying which genetic pathways to test and making sense of experimental data.
- The work demonstrates AI's potential to accelerate discovery in biology by handling literature review and data synthesis tasks that typically consume months of researcher time.
Google DeepMind announced that biologists Omar Abudayyeh and Jonathan Gootenberg have used the company's Co-Scientist AI tool to identify novel genetic factors capable of reversing cellular senescence—a damaged state linked to aging. The researchers conducted large genetic screens by modulating thousands of genes and observing cellular responses, with the goal of pushing cells away from senescence and toward a youthful, functional state in tissues including skin, hair, and muscle.
Co-Scientist contributed in two stages. First, it generated candidate leads by scanning tens of thousands of scientific papers and proposing more than 20 plausible genetic factors worth testing. Laboratory validation confirmed that at least two of these AI-proposed factors successfully drove cells into a younger state with improved overall cellular function. Second, the tool accelerated data interpretation: analyzing screening results alongside scattered scientific literature—a task that typically requires up to six months of researcher effort—was completed in a matter of days.
The researchers characterized the tool's impact as equivalent to having a team of 50 people working within a single day. DeepMind framed this as evidence of AI's capability to support paradigm-shifting discoveries in biology by automating the integration of experimental results with published knowledge, thereby freeing researchers to focus on hypothesis validation and experimental design.
- May 22, 2026 · arXiv cs.AI
New Method Improves LLM Reasoning About Conflicting Beliefs in Complex Social Scenarios
Trust79 - May 20, 2026 · OpenAI — News
OpenAI model resolves 80-year-old discrete geometry conjecture
Trust67 - May 20, 2026 · arXiv cs.AI
Study evaluates how language models interpret personal health records to answer patient questions
Trust74