AI tool Co-Scientist accelerates infectious disease research at Cambridge
At the University of Cambridge, Professor Clare Bryant is employing Google DeepMind's Co-Scientist, an AI tool that operates with multiple agents, to pinpoint molecular switches linked to severe illnesses like sepsis, particularly when pathogens transition from animals to humans. This tool evaluates and ranks hypotheses derived from grant applications and unpublished research, honing in on candidate proteins down to specific amino acids for experimental validation. Bryant’s team is developing cell lines with targeted mutations, aiming to achieve in six months what typically requires two to three years. Co-Scientist integrates published studies and online information to prioritize research inquiries, ensuring that no critical details are overlooked in data-intensive areas, focusing on emerging zoonotic diseases like Ebola, HIV, flu, and Covid-19.
Key facts
- Professor Clare Bryant at the University of Cambridge is using Co-Scientist to hunt for molecular switches causing severe diseases like sepsis.
- Co-Scientist is a multi-agent AI tool developed by Google DeepMind.
- The tool generated and ranked hypotheses from a grant proposal on flu in birds and humans.
- Co-Scientist prioritized a protein connected to signaling pathways Bryant was already interested in.
- Bryant fed unpublished data into Co-Scientist, which remained confidential within the tool.
- The hypotheses moved from candidate proteins down to specific amino acids for experimental focus.
- Bryant's lab is building cell lines with amino acid mutations to test the hypotheses.
- The work is on track to complete in six months instead of the usual two to three years.
Entities
Institutions
- University of Cambridge
- Google DeepMind
Locations
- Cambridge
- Brussels