Tuesday, September 29, 2026 Newsletter Advertise
Breaking
AI

Microsoft Research unveils Quine, an AI world model of biology

Microsoft says its experimental Quine system combines a multimodal biology model with a research harness, and ranked cancer compounds later tested with the Broad Institute.

Microsoft Research unveils Quine, an AI world model of biology. Source: Microsoft

Microsoft Research on September 29, 2026 introduced Quine, which it describes as a research effort to build a multimodal world model of biology paired with an interactive harness linking models, scientific tools, literature and researchers. The company said it has also opened applications for the first cohort of a Quine Fellows program.

What Microsoft says it built

According to Microsoft Research, Quine consists of two components: a world model of biology and a harness that connects that model to orchestration and reasoning models, scientific tools, the literature and the teams of scientists using them.

The company said the world model learns shared representations across biological modalities and scales, including sequence, structure, function, cellular state and imaging data, so that evidence from one modality can inform predictions in another. Microsoft described a world model as a system that can represent the state of a biological system, predict how that state will evolve in response to interventions, and reason over the consequences multiple steps into the future.

Microsoft said learning across these connected representations strengthens rather than dilutes performance, and argued the model "doesn't need to be perfect" as long as it usefully informs experimental design.

The cancer biology example

In collaboration with researchers at the Broad Institute of MIT and Harvard, Microsoft said it used Quine to predict and prioritize thousands of compounds based on their potential to shift pancreatic ductal adenocarcinoma (PDAC) tumor cells between therapeutically relevant states.

In wet-lab studies focused on the transition from classical to basal cell states, the company said Quine's highest-ranked compounds produced the largest intended shifts across experimental assays, and that the process from narrowing the search space to prioritizing candidates for lab validation took just one weekend, potentially saving months of experimental work.

Microsoft said the reverse transition, from basal to classical states, proved more difficult, which it says Quine also predicted. The experiments additionally showed several compounds moving cells toward a distinct third phenotype, an outcome the company said was consistent with Quine's predictions and suggests the PDAC cell-state landscape is richer than a simple classical-basal axis. Some of the strongest effects, Microsoft said, came from compounds with unexpected mechanisms of action.

Access, limits and caveats

Microsoft stressed that Quine is experimental research technology intended only for research, not clinical or medical use, and that its outputs may be incomplete or inaccurate and require review by qualified researchers plus appropriate scientific and experimental validation.

The company said it is taking a deliberate, phased approach to development and access, with initial availability limited to the Quine Fellows program and select research collaborations, supported by ongoing internal review and built-in safeguards. As the technology matures, Microsoft said it expects to expand access through products such as Microsoft Discovery.

The blog post was authored by Nicolo Fusi, VP and Distinguished Scientist, and Jonathan M. Carlson, Vice President, who wrote that Microsoft Research has worked at the intersection of computation and biology for more than two decades.

What to do

  • Scientists who want access can apply to the first Quine Fellows cohort, which Microsoft says is open now; the company says the program gives a cohort of scientists access to the system and a chance to provide scientific feedback.
  • Treat any Quine output as experimental: Microsoft says the system is for research only, not clinical or medical use, and that results need review by qualified researchers plus scientific and experimental validation.
  • Organizations not selected for the fellowship or a collaboration should note Microsoft says wider access is expected later through products such as Microsoft Discovery.
Key facts and where they come from
  • Quine is described as a research effort combining a multimodal world model of biology with an interactive harness.
    Quine (opens in new tab) is a research effort to create a multimodal world model of biology and an interactive harness connecting models, scientific tools, literature, and researchers.
  • Microsoft worked with the Broad Institute to prioritize and validate compounds predicted to shift tumor states.
    we have used this system to prioritize compounds predicted to drive therapeutic tumor-state shifts and validated several top-ranked candidates across multiple wet-lab assays.
  • The compound prioritization work took one weekend, Microsoft said.
    took just one weekend, potentially saving months of experimental work and significant research costs
  • Top-ranked compounds produced the largest intended classical-to-basal shifts in wet-lab assays.
    In wet-lab studies focused on the transition from classical to basal cell states, Quine's highest-ranked compounds produced the largest intended shifts across experimental assays.
  • Quine is for research only and its outputs may be inaccurate.
    Quine is experimental research technology intended only for research, not clinical or medical use, and its outputs may be incomplete or inaccurate
  • Initial access is limited to the Quine Fellows program and selected collaborations.
    with initial availability limited to the Quine Fellows program and select research collaborations
  • Applications for the first Quine Fellows cohort are open.
    Applications for the first Quine Fellows cohort open now

Read the original from Microsoft →

The TechUpscale Brief

The day's cyber, AI and tech news in one short email, every weekday morning. Free. Unsubscribe anytime.

I'm most interested in

More AI