Skip to content

Accelerating the Discovery of Liver Disease Mechanisms

· DeepMind Translated
DeepMind

May 19, 2026 Science

Accelerating the discovery of liver disease mechanisms

Image 37: A medium shot of a man with dark hair and a beard, smiling and looking directly at the camera. He is wearing a light-colored button-up shirt with a dark blue vest over it. His right hand is raised beside his face, with his thumb and index finger pinched together as if making a gesture or explaining something. The background shows a room with a light blue wall, a computer monitor displaying a natural landscape, and white blinds behind a window.

Biomedical research generates vast amounts of information that no scientist could ever fully absorb. Filippo Menolascina, a bioengineer at the University of Edinburgh, is using Co-Scientist to sift through the literature, find overlooked connections, and generate new hypotheses.

His team is focused on a common liver disease: metabolic dysfunction-associated steatohepatitis (MASH). Because MASH involves intertwined biological processes, including liver inflammation and metabolism, developing treatments is challenging, which means single-target drugs are often not enough. This has pushed researchers toward combination therapies, but the number of possible drug pairings is staggering.

Faced with this combinatorial explosion, Menolascina used Co-Scientist to narrow the search. In his hands, Co-Scientist integrated evidence from liver biology and pharmacology, highlighting mechanisms worth prioritizing and flagging candidate combination therapies that his team could test.

In one representative case, Co-Scientist tackled a practical real-world question: why does resmetirom — a recently approved treatment for patients with certain stages of MASH — help only a small fraction of eligible patients? The system proposed a hypothesis identifying the NLRP3 inflammasome as a specific molecular bridge linking inflammation and metabolism, something that had not previously been integrated into a single, actionable explanation. That hypothesis was later experimentally validated, potentially paving the way for a targeted dual therapy.

Co-Scientist is like a rocket booster for scientists, enhancing our ability to identify promising mechanisms. I think we are on the brink of a scientific revolution that will dramatically shorten the iteration cycle needed to achieve breakthroughs

Professor Filippo Menolascina

University of Edinburgh

Co-Scientist: A multi-agent AI companion accelerating research

May 2026 Science

Image 38

Discovering repurposable drugs to combat liver fibrosis

May 2026 Science

Image 39

Integrating a biology toolkit to explore new therapies for ALS

May 2026 Science

Image 40

Opening up new paths in aging research

May 2026 Science

Image 41

Searching for molecular switches behind emerging infectious diseases

May 2026 Science

Image 42

Accelerating the search for gene clues that can reverse cellular aging

May 2026 Science

Image 43