Cure Member Spotlight: Gilead Kedem, MD

Meet Gilead Kedem, MD, CEO and Co-founder of PolyCypher Health, a company using genomics and AI to predict disease risk and enable earlier prevention. Dr. Kedem was also one of the earliest members of the Cure Membership community. We spoke with him about why treating chronic disease risk is the next frontier in prevention, how PolyCypher is moving beyond one-size-fits-all polygenic scores, and his top priorities as the company prepares to launch commercially.

Please tell us a little more about yourself and your team, and how you started on this journey together. 

I’ve spent over two decades in biotech—as a bioinformatician, consultant, and executive— focusing on genomics-based drug discovery and precision medicine. Although I have a background in physics and software, I became interested in health applications following a sudden death in my family. After a 12-year stint in the biotech hotbed of Cambridge, Mass., I became increasingly interested in the potential for genomics to transform healthcare, and I went on to study medicine. In medical school and afterwards, I was struck by how little emphasis is placed on prevention. 

PolyCypher Health came together when Scott Sauers and Edward Messick decided to form a company around their shared background in polygenic scoring, and Edward then brought me on to complete the team. Because Edward was the first person I hired back when I led the R&D Bioinformatics unit at GENEWIZ (now Azenta Life Sciences), we have been close colleagues for over a decade. With complementary backgrounds and a shared vision for advancing precision health and therapeutics, we make a great team of three co-founders. Scott Sauers, COO, is our AI/genomics expert. Edward, CTO, has extensive bioinformatics experience and, while at Genomic Prediction, built the earliest clinical application of polygenic scoring. As CEO, I bring a deep background at the dynamic nexus of technology, biotech, medicine, and business. 

Give us a brief pitch for PolyCypher Health. What do you do, for whom, and what problem are you solving?

PolyCypher’s mission is to predict, prevent, and cure common chronic diseases based on the unique makeup of each individual. Chronic diseases account for the vast majority of sickness, death, and healthcare costs, yet most can be prevented or mitigated. To do so, we are developing a predictive AI platform based on genomics and health data. The platform powers HealthyMe.ai, our genetic test based on a simple cheek swab. Through our platform and genetic test, PolyCypher provides individualized risks for a range of chronic diseases and cancers, tracks health status over time, and provides tailored preventive health guidance. We provide priceless knowledge to extend healthspan and lifespan. 

PolyCypher works with health enterprises, including concierge and longevity clinics, digital health firms, health systems, and insurers, as well as their members. In addition, we work with biopharma firms, applying our core technology to clinical-trial optimization and other use cases involving complex diseases.

What part of prevention did you feel was underserved enough to warrant starting PolyCypher Health, and what early signal convinced you the timing was right?

Chronic diseases—including cardiovascular disease, type 2 diabetes, cognitive decline, and certain cancers—are preventable based on modifiable risk factors. Yet 60% of American adults have at least one chronic disease, leading to almost $4 trillion in healthcare costs. Current healthcare is actually “sick-care”: it is reactive, not proactive, and preventive. However this approach is shifting, driven by consumers' surging interest in longevity, wellness, and preventive health solutions. 

We believe that precision preventive health starts by understanding and treating risk, not disease. However, mainstream healthcare inadequately assesses genetic risk for chronic conditions. Clinical genetics focuses on rare variants, which are informative for rare, monogenic diseases but insufficient to assess risk for chronic conditions. Chronic conditions are complex diseases—caused by an interplay of genetic, lifestyle, and environmental factors—and typically arise from a polygenic background in which many thousands of common genomic variants collectively give rise to risk. For example, over 90% of women with breast cancer do not have the rare BRCA1 or BRCA2 mutations. PolyCypher is transforming healthcare’s inadequate approach to chronic disease by predicting disease risk decades in advance in order to prevent occurrence in the first place.  

Many products stop at “here’s your risk.” How are you turning genetic and longitudinal data into specific, day-to-day actions that people actually follow?

PolyCypher’s vision is to offer a complete solution for preventive health, in which risk prediction is just the first component. Risk assessment starts with our HealthyMe.ai test, based on our advanced polygenic risk score technology. A polygenic risk score measures an individual’s genetic susceptibility to a disease based on thousands to millions of common genomic variants associated with the disease. Next, for superior disease-risk prediction we will integrate electronic health data via our ensemble AI framework.

In addition to personalized risks for a range of chronic diseases and cancers, our report provides preventive health guidance for high-risk diseases, comprising information on lifestyle modification, early screening and early medical intervention. Knowing your outsized risk for a specific disease is an important motivating factor for lifestyle modification, like quitting smoking. For high-risk diseases, early and more frequent screening is an important part of prevention, as is the possibility of preventive medical intervention, for example starting a statin early for high-risk cardiovascular disease. 

For members who subscribe to our HealthyMe.ai portal, we will provide ongoing health tracking to enable members to monitor and reduce disease risk over time. At the enterprise level, PolyCypher will also offer aggregate analytics to allow large healthcare systems to continuously learn about the health of their patient base.

Polygenic risk scores vary by ancestry, calibration, and cohort. How do you address portability, bias, and external validation so results are meaningful across populations?

Polygenic scores are often noisy, biased, or incorrect across people with different genetic ancestry backgrounds. Traditional genetic risk score methods use a “one-size-fits-all” risk curve, which fails when applied to diverse populations. But our model learns a universal adapter for the genetic score, adjusting its predictions based on each person’s specific ancestry so the results apply to everyone, not just the groups represented in the training data. This doesn’t mean the risk estimates are equally accurate for everyone. To account for this, after predicting disease risk, we predict how noisy or inaccurate each prediction is. This allows us to change our estimates based on how confident the model is. For example, someone with a high score but a low confidence might have a medium risk. The last step is to translate scores into lifetime disease risk probabilities, giving doctors and patients a real-world interpretable risk. To validate our model, we use diverse cohorts not seen in the training data, which include people of all ages and ancestries. We measure how fair the score is across populations, the concordance between the probabilities our model outputs and the true disease risk probabilities, as well as the predictive power of the estimate.

Prevention often reaches the already-engaged. Can your solution serve lower-resource settings and diverse genetic backgrounds from day one?

Initially, we will work with concierge and longevity clinics, which are affluent settings where both doctors and patients are highly engaged. This will let us build credibility and iterate quickly based on feedback. Later, we will work with digital health firms, health systems, employers, health insurers, and life insurers. 

We believe that everyone should have access to our cutting-edge precision health services, including lower-resource communities where chronic diseases such as obesity, type 2 diabetes and cardiovascular disease are highly prevalent. The beauty of our business model is that our costs are relatively low, and we have flexibility in pricing models. From the outset, we intend to structure deals with organizations serving lower-resource communities, including those with diverse genetic backgrounds. In addition, whereas polygenic scoring is typically based on biobanks mainly representing people of European ancestry, PolyCypher has developed proprietary technology for superior risk prediction across diverse ancestries.

If you succeed, what changes for consumers, clinicians, and payers in five years?

PolyCypher is creating the future of healthcare, which will be personalized, predictive, and proactive. This future is just around the corner, enabled by AI, the falling cost of genomic sequencing, and the growth of biobanks. As genetics and health data continue to scale, AI models will improve over time, delivering outstanding risk prediction, early detection, and decision-support systems. Ideally, healthcare will evolve into a data-driven, continuously learning system in which incentives align around efficiency, personalized prevention, and long-term value to create a healthier society.

PolyCypher will greatly benefit consumers, clinicians, and payers over the next five years and beyond. Consumers will have greater agency to actively maintain their own health. High-risk individuals will adopt targeted lifestyle changes, earlier screening, and preventive therapies, while low-risk individuals will avoid unnecessary interventions. Clinicians will gain interpretable, actionable insights to provide risk-driven as opposed to symptom-driven care. They will partner with their patients on prevention and early detection. Finally, payers will deploy preventive resources more precisely, intervene earlier for high-risk members, and reduce long-term claims associated with chronic disease. This is a clear win for all stakeholders.

What plans does PolyCypher Health have for the biotech sector?

PolyCypher also works in the biopharma sector, applying our core risk prediction technology to several use cases involving therapeutics R&D for complex diseases. Our first collaboration is with Talon Biomarkers. By combining Talon’s focus on immune system dynamics with PolyCypher’s capabilities in polygenic scoring, we are creating novel diagnostics to monitor inflammatory and oncogenic processes. Because inflammation is a key driver of chronic disease, our collaboration creates multiple business opportunities. 

Another use case involves clinical trial optimization. Clinical trials for chronic diseases are particularly expensive, and polygenic background is an underappreciated complicating factor. By stratifying cohorts based on polygenic risk, we can select participants for clinical trials to reduce costs and increase odds of success. We do so through prognostic enrichment, selecting participants at high risk for a disease, and predictive enrichment, selecting those more likely to have an outsized benefit from the trial intervention.

PolyCypher’s health and biopharma businesses are synergistic, in that learning from each business informs the other. Our biopharma R&D pushes the envelope, and we will gradually incorporate these innovations into our health business. Also, we may reach out to individuals from our health business when selecting participants for clinical trials.

And finally, what are you hoping to gain from Cure’s ecosystem? What’s the one thing that would accelerate your mission right now?

Cure is a special community based in New York City for biotech professionals to come together to network and learn. This kind of ecosystem is particularly important here in New York, where biotech is just one industry in a large, diverse economy. As an entrepreneur leading a startup with enormous growth potential, it’s important for me to both learn from and guide other entrepreneurs on the same journey. 

The Cure is an excellent springboard for PolyCypher to achieve our goals. This is an exciting year for us, because we plan to launch commercially and achieve our first revenue. We have an excellent product, backed by leading technology, a solid business model, and promising market signals. 

Currently, my top priority is to complete our $1 million pre-seed round. We are targeting a range of investors, including angels, high-net-worth individuals, family offices, and VCs. Now is an excellent time to invest in PolyCypher, before we scale, particularly for investors interested in AI, wellness, longevity, digital health, or biotech—since PolyCypher is at the intersection of all these fields.