Skip to main content

Every immune system has a story to tell; the key is knowing how to listen

Close-up of woman working in a medical laboratory

Learning how to listen to immune systems was the challenge that united Microsoft and Seattle-based Adaptive Biotechnologies one year ago. We set out to completely transform the way we diagnose, monitor and treat disease by creating a universal blood test to scan the body for past and present signs of infections, cancers and autoimmune diseases.

The basis of our approach is the T-cell Antigen Map: nature’s mapping of your disease-fighting T-cells to antigens, the signals of disease that they target. If we can learn to decode this information, a small blood sample would yield a readout of the antigens your immune system is monitoring, allowing your immune system to tell its own story.

[msce_cta layout=”image_center” align=”center” linktype=”blue” linkurl=”″ linkscreenreadertext=”Find out more about our journey” linktext=”Find out more about our journey” ][/msce_cta]

We are inviting researchers, biobanks and patient groups around the world to join us on our journey as we immunosequence 25,000 individuals.

Decoding nature’s antigen map requires immunological data of a scale never-before generated, coupled with new algorithmic approaches to modeling how T-cells bind to antigens. To this end, we have been building out the physical lab infrastructure to measure trillions of antigen-binding data points, as well as the cloud and AI infrastructure to model the resulting data, with the aim of making the map clinically actionable.

More important than the physical and digital infrastructure is the integration of the cultures of tech and biotech. Harlan Robins, co-founder and head of innovation at Adaptive, Ryan Emerson, senior director of Antigen Map at Adaptive, and I have been working to build a cross-company, cross-culture team of immunologists, molecular biologists, data scientists and software engineers that sit at Adaptive’s headquarters in Seattle’s Eastlake district.

We are incredibly excited about our progress thus far. Even while we are still building capacity, we have already pushed the state-of-the-art in predicting antigen binding, refined our ability to diagnose a viral infection, and developed a method for using T-cell sequences to accurately estimate the genetic risk associated with many autoimmune diseases. We look forward to sharing our findings in upcoming scientific venues.

A universal diagnostic cannot be achieved alone. And so, we have been working with partners who have deep expertise in specific disease areas to sequence millions of T-cells from thousands of patients.

Recently, Adaptive announced a partnership with the University of Florida to profile over a thousand individuals with (or at risk of) type I diabetes to identify the molecular signals that could enable early diagnosis and treatment. Similar collaborations are underway with investigators at a number of institutions, including the Benaroya Research Institute at Virginia Mason, University of Colorado Anschutz Medical Campus, and The Fred Hutchinson Cancer Research Center to profile individuals with different diseases.

Today we are excited to announce the global expansion of this program.

We are focusing our initial efforts on five specific diseases based on unmet clinical need and our understanding of the underlying immunology: type 1 diabetes, celiac disease, ovarian cancer, pancreatic cancer, and Lyme disease.

We are inviting researchers, biobanks and patient groups around the world to join us on our journey. Our aim is to sequence the T-cell repertoires of 25,000 individuals affected by one of the five specific diseases we are initially pursuing based on unmet clinical need and our understanding of the underlying immunology: type 1 diabetes, celiac disease, ovarian cancer, pancreatic cancer, and Lyme disease. These diseases represent some of the different roles T-cells play in controlling or causing autoimmune diseases, cancers and infections.

Concurrently, we are focusing our Antigen Map efforts on known and suspected antigens for each of these target diseases. Although T-cell sequences can be treated as just another marker to mine for associations with disease, the underlying biology of T-cells tells us that linking specific T-cells to specific antigens in the context of specific diseases will provide key insights into the underlying causal mechanisms of disease.

For example, knowing the optimal T-cells for your specific cancer mutations would enable far more precision in prescribing personalized immuno-oncology interventions that seek to activate or engineer cancer-specific T-cells. Conversely, understanding the common antigens that lead to the destruction of insulin-producing cells in type I diabetes would provide much-needed focus to efforts of designing tolerance-inducing vaccines that could halt or prevent such autoimmune disorders.

So, while decoding the Antigen Map is the basis of our long-term aim of a universal diagnostic, it will also provide a new platform for designing targeted immunotherapeutics and vaccines. For this reason, both companies will continue to collaborate with current and future pharmaceutical partners to ensure that the developments we make toward better diagnostics will also enable transformational treatment options.

At Microsoft, we believe that the convergence of biotechnology and hyperscale machine learning could fundamentally alter the science and practice of medicine. We are incredibly humbled to work with partners like Adaptive to leverage AI and the cloud as we work to make this convergence a reality.