【Edgewater Corporation】AI Analysis of Immune and Genetic Data with RIKEN. The Medical Revolution of the "Mail-In Easy Collection Model" for Real-World Data (RWD) That Pharmaceutical Companies Covet
You Can't Prevent Without Predicting: Integrated AI Analysis of Immune, Genetic, and Environmental Factors in Partnership with RIKEN
Nice to meet you all. My name is Masahiko Fukuzawa, Representative Director of Edgewater Corporation. Today, I would like to introduce our "Prediction and Prevention Healthcare Data Platform."
While it is often said in society that "preventing diseases is important," we take it a step further. We established this business based on the strong conviction that "unless we predict the diseases a person is destined to contract in the future with high accuracy, we cannot implement concrete prevention."
We design our data analysis model based on the theory proposed by British scientist Francis Galton: "Diseases develop due to a combination of genetic and environmental factors contributing in various proportions."
Specifically, we integrate three types of data:
- Genetic factors (DNA data that never changes, obtained via genetic testing)
- Environmental factors (records of daily habits such as diet and exercise)
- Current health status (our unique immune testing deciphered from blood samples)
By turning these into big data and analyzing them with AI, we predict individuals' future immune states and specific disease risks.
In conventional medicine and healthcare, because integrated immune, genetic, and lifestyle data does not exist, predicting onset was difficult, and treatment was only possible after falling ill. We have been steadily collecting this data for years. By enabling prediction, we realize "complete preventive medicine" that either prevents the onset of disease entirely or keeps it extremely mild.
The core of this business is the "accuracy of the predictive model."
No matter how advanced AI and big data analysis technologies become, the phenomenon of disease is extremely complex. Therefore, rather than relying solely on AI technology, we chose to collaborate deeply with the advanced medical academic field. Currently, we have a joint research agreement with the autoimmune disease research team at "RIKEN," Japan's leading research institution (the team led by Dr. Kazuhiko Yamamoto).
They are a world-class team that has researched diseases for years on the twin axes of "immunity" and "genetics," leading to the elucidation of causes and the creation of treatments. RIKEN's goals and our concept of "fusing prediction and prevention" aligned perfectly, leading to the launch of this joint research.
The "Human Raw Data" That Pharmaceutical Companies Crave: A Patented Model Bridging the Gap Between Mice and Humans
To make the business successful, the monetization model is extremely important. In the healthcare and medical industry, the entity with the strongest and most sustainable purchasing power is undoubtedly the "pharmaceutical company."
Today, pharmaceutical companies face an extremely large challenge.
As stated urgently in the latest report by the Japan Pharmaceutical Manufacturers Association (JPMA), Japanese pharmaceutical companies are smaller in scale compared to global mega-pharmaceuticals. To win the competition, the utilization of "Medical Information Databases (Real-World Data: RWD)" is indispensable.
To advance R&D, new drug discovery, and clinical monitoring, everyday medical records (clinical charts) are completely insufficient. Pharmaceutical companies crave disease-specific, highly detailed biological profile data, such as genome, immune (proteome), omics, and microbiome data.
However, researchers at pharmaceutical companies face a critical bottleneck. Because they are not medical doctors, they are legally prohibited from directly contacting patients (humans) to collect data. Consequently, they must conduct their basic research using laboratory animals such as mice and rats.
Yet, the immune systems of mice and humans differ significantly. There are numerous cases where a drug that succeeded in mice failed in humans. Pharmaceutical companies desperately want "detailed human immune data."
We have built a unique collection model to clear this challenge and have obtained a patent for it.
In our model, without going through medical institutions (hospitals), we collect data directly from individuals using home-use "easy mail-in testing kits" (genetic testing via saliva and immune testing via micro-blood self-sampling).
In return for providing users with a "future disease prediction report" for free, we obtain consent directly from them for the secondary and research use of their data by pharmaceutical companies. This creates a unique B2B marketplace that supplies pharmaceutical companies with the "consented, continuous human immune and genetic data" they want most, both cheaply and at scale.
As evidence of past data transactions, in 2018, the US genetic testing company "23andMe" provided its accumulated consented genetic data of 5 million people to GlaxoSmithKline (GSK) for approximately 45 billion yen (at the time). GSK built over 40 drug pipelines and contributed to Parkinson's disease treatments based on this data. For pharmaceutical companies, such human raw data assets are something they want to acquire even at the cost of tens of billions of yen.
Turning a Once-in-a-Lifetime Genetic Test into an Annual Repeat: Standardizing and Exploding in the US in Partnership with 23andMe
Our growth strategy is to deploy this framework solidly in Japan and then explode in the "US market," where the market size is 100 times larger.
The US-based "23andMe," which I mentioned as an example, currently has an overwhelming customer base of over 15 million people, but it faces a fatal business weakness.
Because a genetic test (DNA test) is taken only "once in a lifetime," once they sell the test, they cannot generate recurring revenue from that user. Consequently, their customer acquisition cost (CAC) has soared, leading to deteriorating profitability.
Therefore, our strategy is to "collaborate" with these massive genetic testing companies in the US.
We will offer our patented "annual periodic immune testing (blood data)" to their existing base of 15 million members.
By overlaying dynamic, annually changing immune data onto static, once-in-a-lifetime genetic data, the accuracy of disease prediction increases dramatically. For genetic testing companies, this establishes a "recurring revenue business model," and for us, it instantaneously secures tens of millions of customer profiles and RWD for pharmaceutical companies. Even if only 0.5% of their members take our test, it gathers hundreds of thousands of profiles, launching a data business worth tens of billions of yen in an instant.
Armed with this patented model, we will accelerate new drug discovery for pharmaceutical companies and provide individuals with "accurate disease prediction and personalized prevention before falling ill." Through this, we will build a health infrastructure that fundamentally reduces medical expenses in Japan and globally.
For investors who aim to transform healthcare through data and science, and for companies wishing to collaborate, let us build this grand future together. Thank you very much.
Q&A and Feedback
Commentator (Mr. Fukutani): Thank you, Mr. Fukuzawa. The deep pain points of the medical and pharmaceutical industries and your unique data platform model are very clear.
While "pre-symptomatic state" and "prevention" before falling ill are highly popular fields, the greatest bottleneck is "how to collect data from ordinary individuals." How do you collect genetic and blood data from individuals who find it troublesome to visit hospitals? Please tell us about the specific collection process.
Mr. Fukuzawa: Thank you for the question. Indeed, the "data collection channel" is the core of our business.
Conventional medical tests required visiting a hospital and waiting a long time for blood sampling, which was a high psychological hurdle. We deliver "easy mail-in testing kits" (self-sampling kits for micro-blood from the fingertip) and saliva-based genetic testing kits directly to the user's home. Since it is completed at home in a few minutes without visiting a hospital, we can lower the hurdle of data collection to the limit.
Furthermore, we utilize the fact that pharmaceutical companies have research needs for specific diseases (e.g., autoimmune diseases) to attract users. We recruit individuals who are at risk or interested in those diseases online and provide "free testing and AI-driven future disease prediction feedback" funded by pharmaceutical companies. Because users see a clear benefit in "knowing their future health risks for free," they willingly undergo testing and consent to the secondary use of their data.
Mr. Fukutani: I see. You provide users with free high-precision health checks, and the source of funding is paid by pharmaceutical companies seeking the data. It is a beautiful and rational ecosystem.
However, regarding why pharmaceutical companies pay such huge sums of money for your data, could you explain the difference between your data and the clinical trials they conduct themselves?
Mr. Fukuzawa: For pharmaceutical companies, the most critical process is "clinical trials" to verify the efficacy of new drugs.
However, because the researchers themselves are not medical doctors, they are legally prohibited from directly examining patients to gather data outside the highly restricted framework of clinical trials. Consequently, they have no choice but to use mice and rats in basic research.
Yet, the immune systems of mice and humans differ significantly. While pharmaceutical companies crave human data, gathering it outside clinical trials through medical institutions requires massive costs and ethical/legal procedures.
Our model operates outside medical institutions, where individuals voluntarily undergo testing as a free healthcare service and consent to the secondary use of their data. This allows us to accumulate RWD encompassing detailed human biological data very cheaply and safely. This collection process is patented, forming a strong barrier that competitors cannot easily duplicate.
Mr. Fukutani: Thank you. You have bypassed the structural dilemma of pharmaceutical companies' inability to access human data directly through patented easy mail-in kits and free user returns.
Also, regarding the US expansion, I found the specific image of collaborating with genetic testing companies like 23andMe highly interesting. Could you explain in more detail how you complement their monetization issues?
Mr. Fukuzawa: Yes. 23andMe gathered 5 million genetic profiles and achieved major results, such as providing data to GSK for 45 billion yen. However, their model is a one-time transaction. Because genes do not change, they cannot sell tests every year. As a result, they cannot repeat sales, facing difficult management as new user acquisition slows down.
To them, we propose our "annual immune testing (blood data)."
By overlaying dynamic immune data onto static genetic data, the accuracy of disease prediction rises dramatically. For genetic testing companies, they can add a "recurring subscription model" to their existing base of 15 million members, stabilizing their management. For us, we do not need to gather tens of millions of customers from scratch; we can instantly capture the US market by riding on their membership base. This mutual complement is our greatest winning strategy in the US.
Mr. Fukutani: That is highly convincing. Multiplying static genetic data and dynamic immune data to raise accuracy, and changing one-time transactions into recurring models. It is a wonderful strategy that maximizes the differences between the Japanese and US markets and the value of partnership. I look forward to your global expansion.
Mr. Fukuzawa: Thank you very much. We will continue to drive the business to contribute to the health of people worldwide.