1. Skip to Menu
  2. Skip to Content
  3. Skip to Footer

We need a CERN for genomics


Emmanouil Dermitzakis is currently the Louis-Jeantet Professor of Genetics at the University of Geneva. From 2004-2009, he was a group leader at the Wellcome Trust Sanger Institute in Cambridge. His research focuses on the genetic basis of cellular phenotypes and complex traits. In an interview with Thomas Lemberger, chief editor of the journal Molecular Systems Biology, he talks about the functional use of population genomics, disease biology and the power and challenges of large-scale sequencing projects.


Your research focuses on understanding the genomic basis of human variation. What is your general approach?

Studying variation in humans touches both basic science and medically related questions dealing with disease, treatment and the understanding of disease in general. Our approach is from a basic perspective. We are trying to understand how genetic variation in general – meaning all the variants in the human population – has an impact on cellular processes.


Conducting genetic experiments is possible in mice but impossible in humans. So what is the critical advantage of studying genomic variants in humans at the level of entire populations?

The advantage is not as much scientific as it is social. I think the interest in understanding medically related questions has opened up a lot of doors both on the hospital side as well as at the level of the individual citizen. Sampling tissues from individuals in a non-invasive way is more and more accepted. We can look at cells and generate cell lines, which serve as models to do experiments on. In fac


How do you address the often complex mechanisms that lead to a disease?

There are two aspects to this issue: how you get a particular disease and how you approach its treatment are two different things. One example is diabetes that results from loss of control of glycemia. The causes of the disease are probably longterm. It is a combination of genetics and environment. You cannot just reverse the causes and treat the disease. Because if you could, you would just lose weight and be fine again. Understanding the causes and informing people how to prevent a disease – called preventive medicine – is one thing. But treating a patient who lost the equilibrium is a whole different approach.


Why is it important to sequence more and more individual human genomes?

We should sequence everybody. In fact, people should be sequenced at birth. Not because this will predict a disease. The reason is that each one of us is going to have something that is rare or even private. To better understand a rare condition it is always helpful if you compare to other patients without that mutation. It will also improve our understanding of the impact of the DNA sequence on particular phenotypes. Some people might be able to control it just by changing their behaviour as opposed to drug treatment. I think everybody should be sequenced because it is a fundamental thing. This is similar to looking at yourself in a mirror, weighing yourself or measuring your height. It would be a great benefit to humanity if we reached the level where we can share DNA information without being afraid that somebody will take advantage of it. It might be difficult but we should aim for that.


Sequencing has become very cheap, which explains why we can sequence more and more people. Is it going hand in hand with similar efforts to measure molecular and clinical phenotypes?

Phenotyping of individuals in a medical setting, so called clinical phenotypes, and the recording of those phenotypes in an electronic health record in an organised fashion is not very advanced. It is disappointing that now that the genome is available we have to wait for the electronic health record. It would have been great if those records were already organised. It is late to do it now because every hospital has its own system and in addition the national health systems are often totally different.


Speaking about the scientific process in genomics: research in genomics often involves large consortia. Do you think we need a CERNlike structure for genomics?

We need large consortia because the major efforts need to be done collectively – compiling the data, deciding on standards, organising and sharing available data. Establishing a homogenous way to do all these things has helped genomics very much. On the other hand, innovation will not come from consortia. Consortia are just helping to build a base line that people can step on and move on to discover interesting things. We need a mix of both. In my opinion we do need a CERN for genomics for two reasons. First of all we need structures in Europe that will influence the European setting. What we also need is a place to meet. Because CERN is not just about doing research. It is also a meeting point for scientists. EMBL has fulfilled this role for molecular biology. Genomics is a new revolution and we need to build something similar now.


Is it important to share findings and data prior to publication, posting the manuscripts on bioRxiv (bio-archive) for example?

I am split on this. My group has always been very open about announcing things at conferences and in talks on unpublished data. I really like sharing data, information and discovery. But I am worried about the gossip. Publication, with all its caveats and its disadvantages, is an organised way of sharing information. You cannot start sharing findings and data in an aggressive way. The community needs to mature first. And we need to be more organised about sharing information, particularly publication.


This email address is being protected from spambots. You need JavaScript enabled to view it.


Tilmann KießlingTilmann Kießling
Head, Communications
T. + 49 160 9019 3839