Microsoft Research Debuts Quine, a Multimodal World Model of Biology

Microsoft Research on September 29, 2026, introduced Quine, a research effort built around a multimodal world model of biology and an interactive harness that links the model to orchestration and reasoning models, scientific tools, the research literature, the wet lab, and the teams of scientists using them.

With the announcement, Microsoft opened applications for the first Quine Fellows cohort and reported that, working with researchers at the Broad Institute of MIT and Harvard, it has used the system to rank compounds predicted to produce therapeutic shifts in tumor cell state, with several top-ranked candidates validated across multiple wet-lab assays.

The announcement is authored by Nicolo Fusi, VP and distinguished scientist, and Jonathan M. Carlson, vice president.

A World Model Built Across Biological Scales

Microsoft defines the world model it is pursuing as one that represents the state of a biological system, forecasts how that state would change under an intervention, and reasons through the downstream consequences several steps ahead. The company says such a system would not replace experimentation; instead, it would let researchers weigh and rank candidate directions computationally before committing limited laboratory resources.

Quine’s core model is trained to learn representations shared across biological scales and data types, among them sequence, structure, function, cellular state, and imaging. Its training modalities span genomics, proteins, chemistry, RNA and cell state, and bioimaging. Microsoft reports that joint training across these modalities strengthens rather than dilutes performance: evidence from one modality can inform predictions in another, capturing relationships the company says could stay siloed or be lost if separate single-domain specialist models were orchestrated instead.

In Microsoft’s description, the system runs as a cycle that starts and ends with the scientist: a question yields proposals, proposals turn into experiments worth running, and the resulting measurements refine both the next question and the model itself. The model need not be perfect, the company says, adding that it will never perfectly capture biology; it needs only to guide experimental design usefully.

The Quine project page describes three components developed in tandem: the model itself, built to connect evidence across proteins, cells, tissues, and genomes; a tool layer meant to help researchers decompose questions into steps, apply models and scientific tools, weigh evidence, and revise plans; and a research program in which Microsoft scientists, fellows, collaborators, and experimental partners test the system against real biological questions. The page dates the research lines feeding into Quine between 2021 and 2026, covering proteins, the regulatory genome, the tumor microenvironment, tissue pathology, microscopy, and biological data.

Microsoft Research says it has spent more than two decades working where computation meets biology, across immunology, virology, genomics, biomedical imaging, cell biology, and protein engineering, and that Quine has become central to how its teams pursue discovery in areas including cancer biology, protein engineering, genomics, and bioimaging.

Wet-Lab Validation in Pancreatic Cancer

The announcement grounds the system in a specific case: pancreatic ductal adenocarcinoma (PDAC), which Microsoft describes as the most common form of pancreatic cancer and among the hardest to treat. With Broad Institute researchers, the company has developed and applied patient-derived ex vivo models over years to test a longstanding hypothesis: that how a tumor behaves and responds to drugs depends on its transcriptional cell state as well as its genetics. PDAC tumor cells can sit in different cellular states linked to treatment response.

That collaboration sits within Project Ex Vivo, a joint cancer research effort between Microsoft and the Broad Institute with support from the Dana-Farber Cancer Institute, aimed at defining, engineering, and targeting cell states in cancer. Its initial PDAC work examined the therapeutic significance of cell state and its microenvironmental drivers.

Microsoft reports using Quine to score and rank thousands of compounds for their potential to move tumor cells between states relevant to therapy. In wet-lab studies centered on the shift from classical to basal cell states, the company says, the compounds Quine ranked highest moved cells furthest in the intended direction across the assays, and several of the strongest effects involved compounds with unexpected mechanisms of action, results Microsoft describes as an early indication that AI can surface opportunities for drug repurposing and discovery.

The process, from narrowing the compound search space to selecting a short list of candidates for laboratory validation, took a single weekend, Microsoft says, a pace it describes as potentially saving months of experimental work and considerable research cost.

The reverse shift, from basal back to classical, proved harder, an outcome the company says matched Quine’s prediction that available compounds would produce weaker effects in that direction. The experiments also confirmed a separate Quine prediction that several compounds would repeatedly push cells toward a distinct third phenotype, which Microsoft says suggests the pancreatic cancer cell-state landscape is more complex than a simple classical-basal axis. Next steps in that work include integrating new RNA datasets and tasks, sharpening state-transition predictions, and adding calibrated confidence estimates so scientists can rank the strongest hypotheses for wet-lab testing.

Fellowship Terms, Safeguards, and Phased Access

Microsoft states that Quine is experimental research technology, intended for research only and not for clinical or medical use, and that its outputs can be incomplete or inaccurate, requiring review by qualified researchers and validation through appropriate scientific and experimental work. Access is being rolled out under what the company calls a deliberate, phased approach: initial availability is restricted to the Quine Fellows program and select research collaborations, with ongoing internal review and built-in safeguards. Microsoft says it expects to widen access later through products such as Microsoft Discovery as the technology matures.

The fellowship runs 16 weeks, includes financial support, and is hosted by Microsoft Research in Cambridge, Massachusetts. It is open to PhD candidates, postdocs, research scientists, and academic or independent researchers. Participants will work alongside Microsoft researchers with access to Quine and computing resources, and some projects may also receive experimental support. Listed areas of interest include protein design and engineering; enzyme design, discovery, and optimization; genetic and chemical perturbation of cell state; early-stage therapeutic research in under-resourced disease areas, including hypothesis and compound prioritization; and adjacent problems where faster iteration between design and experiment could change the outcome.

The fellowship page cautions that applying guarantees neither selection nor access, and that participation and support depend on eligibility, selection, capacity, and written fellowship terms. Applications for the first cohort opened September 29, 2026, and run through November 2, 2026; the fellowship itself is scheduled for June 7, 2027, through September 24, 2027.