BioHUG biobank tops 10,000 DNA samples

The BioHUG genetic database aims to store DNA from 20,000 people by 2027, making it available to selected scientific research projects.

Illustration of DNA sequencing. Image: Adobstock

Established in February 2025, the BioHUG biobank recently recorded its 10,000th DNA sample collected from patients at the Geneva University Hospitals (HUG). This symbolic milestone puts the project at the halfway point towards its target of 20,000 samples — a goal expected to be reached by next year, which will make BioHUG one of the most significant hospital biobanks in Switzerland, comparable to those already operating in the United States and Scandinavia. This genetic dataset represents an invaluable resource for medical research, particularly in the field of personalised medicine. Timothy Frayling, Professor in the Department of Genetic Medicine and Development (Faculty of Medicine) and BioHUG’s Scientific Director, explains.

What is the purpose of a biobank like BioHUG?

This biobank is dedicated to disease research. It brings together samples from patients who have passed through HUG and signed the “general consent” form. This authorises scientists to reuse those samples, along with the associated health data — past, present and future — without needing to contact patients again each time a new use arises. As this is sensitive data, the information is encoded in a fairly complex way to prevent identification of the individuals concerned, while still enabling scientific studies to be conducted. Consent is voluntary and remains valid indefinitely, or until withdrawal, which is possible at any time and without justification. For BioHUG, we are collecting DNA from patients who signed the general consent from 2017 onwards and from blood samples taken since February 2025. We have already collected 10,000 samples and our aim, within the current project funding, is to reach 20,000. That said, 150,000 people at HUG have now signed the general consent, so the potential for growth is considerable.

Does Geneva show a particularly high rate of general consent signatures?

Yes, and that is the case across all major urban centres in Switzerland. When I arrived in Geneva two and a half years ago, I was struck by how remarkably well Switzerland has done in disease research. No fewer than 800,000 people — roughly 10% of the population — have agreed for their medical data and residual samples to be reused for research purposes. No other country has reached that percentage, apart from Finland and Estonia, which are both smaller than Switzerland. The electronic medical records going back to 2000 and continuing through to end of life for hundreds of thousands of people represent a remarkable dataset and, crucially, an extraordinary opportunity for Switzerland to consolidate its position as a leading country in biomedical research.

Why is this data so important, and why is such a large volume needed?

It offers a new way of studying disease. The traditional method — which remains very important — is expensive in both time and money. It involves voluntarily enrolling people with a specific condition into a study, recruiting them, and measuring many aspects of their health, including new blood draws. Participants are then followed for sometimes two or three years to observe the effects of treatments and disease progression. The entire process has to be repeated for every condition and every new research question, which is why these studies are generally limited to a few hundred patients. It is also often difficult to recruit people without the condition being studied — the controls — for comparative purposes.

The biobank approach is to create a single very large study — say, 100,000 hospital patients — and analyse their medical data and biological samples over decades. This allows us to study the causes and consequences of diseases over long time periods, and to include people who do not have time to take part in traditional research studies, since we are simply reusing their coded data. It is also an extremely powerful way of studying people with multiple conditions — multi-morbidity — a phenomenon that is increasingly common in an ageing population. We can, moreover, study people before they develop a disease, since many of them undergo routine health checks at the start of their adult lives.

Do you have examples of research projects?

One of our projects is to assess the risk of heart disease in people with autoimmune conditions. Those living with type 1 diabetes, rheumatoid arthritis, lupus and other such conditions face a higher risk of developing heart disease, and the aim is to measure that risk more precisely. The challenge is that these conditions are — fortunately — rare: they affect only around 4% of the population, and of those, only about 10% will develop heart disease over a ten-year period. In other words, to learn more, we need to study 0.4% of the population, which requires a very large dataset.

Another example relates directly to diabetes. Around 10% of people whose ancestors lived in tropical regions of the world — in Africa and parts of Asia — carry a genetic variant that protects against malaria but which, at the same time, affects how diabetes is measured. This genetic variant shortens the lifespan of red blood cells, which means that the blood sugar level measured on red blood cells — glycated haemoglobin, or HbA1c, a key value in the diagnosis and management of diabetes — appears lower than it actually is. Recent studies have shown that these individuals face a higher risk of diabetes-related complications because their diagnosis is delayed by four years. At BioHUG, we hope to contribute to discoveries of this kind, so that further studies can then examine their direct application for patient benefit. That is precisely the aim of our current project, CALIBRATE.

What does CALIBRATE involve?

When a physician orders a blood test, the result is compared against a single reference range. In the case of diabetes, for example, a healthy person’s blood sugar is estimated to fall between X and Y. The same applies to kidney function, assessed through other specific biomarkers measured in the blood. These ranges are the same for everyone, regardless of age, background or individual biology. However, what counts as “normal” can vary from one person to another. The body tends to maintain a number of its physiological characteristics around a set point that is its. Body temperature in most people is around 37°C, but that is an average: 38°C can be a fever for one person and entirely normal for another. The aim of the CALIBRATE project is to calculate individual reference values for each person based on their genetic background — their personal “norm”, so to speak — and to examine whether using these personalised reference values helps physicians to detect disease earlier, assess risks more accurately, and better understand the extent to which these personal reference values are shaped by our genes.

Media