This month, we interview Dennis van ‘t Ent: associate professor at the department of Biological Psychology at the Vrije Universiteit Amsterdam
Please tell us about your background and how you became an imaging genetics researcher.
I studied physics and astronomy, but I’ve always had an interest in medicine. During the end of my physics studies, I specialized in medical physics and did an internship that involved human studies. I started by doing eye-movement recordings because they are an important window into the brain. After the internship, I did a PhD in Rotterdam on EEG recordings, with a focus on cognition. After that, I returned to Amsterdam to do a postdoc at the VU Medical Center, where I worked at the center for magnetoencephalography, which was a relatively new technique back then. I currently work as an associate professor at the Netherlands Twin Register at the VU, where I teach a lot and focus again on neuroimaging now mainly using MRI scans.
For those in our audience who may be less familiar with imaging genetics, could you explain the basics of this field?
Imaging genetics can be broadly defined as the study of how genes influence our brains. On the one hand, we want to know what genes are relevant and how they affect the brain, and on the other hand, we are interested in the effect on behavior. So we are looking from genes to the brain, to behavior. Besides the genetic contribution, the environment also plays a large role in shaping both the brain and behavior. With twin data, we can tease out the genetic and environmental contributions.
In genetic research, large samples are required to find robust effects. Do you need similar sample sizes for imaging genetics and how attainable is it to get such large samples with MRI data? Imaging genetics generally also needs large sample sizes, although the exact number depends on multiple factors. Older studies used to have smaller samples, but recent research has shown that we do need larger samples for reliable findings. One way to accommodate this is to form large consortia, such as ENIGMA, that pool MRI data from many labs, which has been an important development in the field. We can also make use of large biobanks like the UK Biobank. However, the required sample size depends on the trait you are investigating. It has been shown that for some variables, such as adult age, you need at least 300 people to see an effect related to the brain, but for other traits, you need more than thousands of individuals. It furthermore depends on which brain area you are looking at, as effect sizes can be larger for some than for other regions. Additionally, if you only focus on one region, you will also have a lower multiple comparison problem, in which case you would need fewer individuals.
Is the field theory-driven then, if you can preselect certain brain regions?
It depends on the research question. The whole brain is often included, but if you are interested in for instance executive function, you may focus on the frontal lobe. For emotion, you may study the amygdala.
Are brain imaging traits mediating the effects of variants on traits? Can they provide more insight into biology?
Yes, but this again really depends on the trait you look at. Take substance abuse for instance, where you may expect that genes affect the development of the reward system in the brain, such that you are more sensitive to substance use. But there is also the issue of reverse causality, where people who have a genetic predisposition for substance use will use alcohol or smoke frequently, which in turn affects the brain. We need to tease those effects apart. This is difficult of course, but longitudinal or Mendelian Randomization studies may be useful here.
You are a member of the GENE Amsterdam Scientific Outreach Committee, can you tell us what this committee is currently doing?
We are involved in providing news content on the GENE amsterdam website and social media and we are also working to obtain funding for our activities. We are currently preparing a funding proposal for outreach to high school students in the Netherlands about genetics and its relationship to society.
You do a lot of teaching, what do you like most about it?
There are many things. But what I enjoy the most is the interaction with the students and getting them motivated. For example, I teach biological psychology and some of my students have limited backgrounds in biology and have low expectations of the course. But often they become very enthusiastic during the course; it feels good having played a role in that.
Is it difficult to combine teaching with research?
Yes, time is always an issue. I teach in blocks, and in some months there’s barely any time for research at all. That can be very frustrating, as the science never stops. On the other hand, teaching is very important and should not always take the second rank to research activities. One of a university’s main goals is to deliver teaching. It feels like this has been forgotten for a long time and therefore I think the recognition and rewards initiative at universities system is such an important development.
Do you have any favorite brain regions?
Oh, difficult question. It’d probably be the frontal lobe, as it’s related to my PhD work on action monitoring. It’s like the control center of the brain.
If you weren’t a geneticist, what other field might you have pursued and why?
I would very much have liked to be the goalkeeper of Ajax Amsterdam. But to be more realistic, I am fascinated by the laws of physics and observing them in the universe – a remnant passion from my time studying astronomy. I could have seen myself being an astronomer in the deserts of Chile, looking up at the sky and doing calculations on the stars and the orbits of the planets..
Opportunity to join our Research Meeting Committee
Is your new years resolution to become more involved? To build your CV? Or perhaps to get to know more scientists within your field? This is the opportunity you have been waiting for!
We are looking for a volunteer to join Wonu Akingbuwa the research meeting committee. Joining this committee means helping organize the monthly GENE amsterdam research meetings: looking for speakers, inviting the community, and making sure we have an online or live location. Interested? Please contact Wonu at: firstname.lastname@example.org.
Interested in helping out, but not neccesarily with organizing the research meetings? We are also looking for people to help out with social media and organizing the annual day. Please contact Margot van de Weijer for more information at email@example.com.
Can you tell us about your background and how you became both a clinical geneticist and researcher in complex trait genetics?
I went to Erasmus Medical Center where I completed my medical degree in 2014. Interestingly, my medical training included very little education in genetics, which is surprising given its growing importance in healthcare. After graduating, I pursued my Ph.D. in the Department of Complex Trait Genetics led by Danielle Posthuma, where my focus was on polygenic risk scores for psychiatric disorders and brain imaging. After obtaining my Ph.D. I started my clinical training, which I will finish next year, and will start working as a clinical geneticist at the Amsterdam UMC (location AMC). Currently, my clinical work is mainly focused on monogenic diseases, such as heritable forms of Alzheimer’s disease and dementia, movement disorders, and developmental problems in children. My research is mainly focused on complex trait genetics in the broadest sense.
Is there any significant overlap between your clinical and academic work?
The overlap is currently minimal, unfortunately. That’s because, in monogenic disease genetics, there is hardly any involvement of common or complex genetics in genetic diagnostics. However, this is slowly changing with the development of polygenic risk scores. Recent papers (e.g. Khera et al. (2018)) have shown that for some conditions, being in a high percentile of polygenic risk can be compared to a relative risk of a monogenic mutation that we test for in the clinic. People are now thinking about ways to bring polygenic scores to the clinics. To provide a potential example I encountered in my own work: We have an expertise center for genetic obesity, where individuals are tested for monogenic forms of obesity. A subset of patients have a very high polygenic score and do not have monogenic mutations for obesity. Whereas in the lower polygenic score segment, we find many more rare monogenic mutations. I think that’s a way we can use polygenic risk scores in the future, by informing us in which individuals a monogenic cause might be found and aiming our genetic sequencing mostly on that group.
Do you see a role for a clinical geneticist in the field of psychiatric disorders as well?
Yes, I do. We often see patients who have multiple family members with, for example, schizophrenia who want to know their children’s risk of that disorder, which is very difficult to estimate based on the literature.
It would be great to improve our estimates using genetic testing. There are recently discovered rare mutations that increase the relative risk of schizophrenia by 50 times. It would be useful to incorporate this kind of information into the clinic, and maybe combine it with polygenic risk scores to further improve our risk estimates.
What does your research focus on?
I am currently working on the relationship between sleep problems and dementia, which have an epidemiological correlation with each other, but we’re still not sure whether one causes the one or the other way around, or whether there are genes that could explain this correlation. To tackle this, I am looking into the genetics of brain aging. With machine learning, we can estimate someone’s age from their MRI brain scans with an average error of about three years. We did a GWAS on the gap between someone’s actual age and their predicted age based on their brain scans. The associated genetic variants were not strongly related to insomnia, but we found many significant genetic correlations with psychiatric disorders, mainly anxiety-related disorders. You can imagine that if you are more anxious, you may have higher cortisol levels or a less optimal healthy lifestyle which could explain this finding.
How do you balance your time between your clinical and academic work? When working in the clinic, it is sometimes difficult to switch gears to focus on research. Alternating weekly between both helps but clinical work always goes on, so you cannot separate it that easily. If a patient needs to be called about new test results, this always comes first. While it can be tricky, it also has benefits. Clinical work can be quite overwhelming, but it does provide you with a lot of diversity that research doesn’t, because you’re working with new people every day, and every patient contact is different. Clinical work brings you a lot, but there is a commitment and offer that you need to be able to make. It also makes it easier to come up with clinically relevant research questions and fill a need.
What does a typical day in the clinic look like? This morning we started with patient rounds, which are multidisciplinary discussions with other specialists. Then I saw three new patients who were referred to our clinic. We have plenty of time for each new patient, which is relatively rare in a hospital. This time is necessary since an important part of the intake is to look at the family history and draw a family tree with the patient to get a first idea about the inheritance pattern of the disease. Afterward, I had a research discussion about a polygenic risk score project. But the bulk of my days (like many other clinical specialties) are taken up by the administration.
If you could have dinner with any scientist (living or dead), who would it be and why?
There is an influential Dutch geneticist who you hardly hear about. His name is Hugo de Vries, a contemporary of Mendel’s who coined the term mutation. Similar to Mendel, he’s done a lot of plant experiments and was an important discoverer of his work. Can you imagine telling him about GWAS, that would probably blow his mind. I think it is time to start thinking about developing a Hugo de Vries prize for outstanding scientists in the field!
November 28, the second annual GENE Amsterdam Day was held in the Doelenzaal in Amsterdam. We reflect back on a wonderful day full of interesting talks and discussions.
We started the day with a recap by Margot van de Weijer on the work done by the different GENE Amsterdam committees in the past year, followed by our the first symposium Mental Health and Education. We kicked off the symposium with a talk from Bruno Sauce Silva (department of Biology Psychology, VU Amsterdam) on his work studying GxE in an educational context in the US. Next, Rick Jansen (department of Psychiatry, AUMC, location VUmc) presented on the metabolome-wide signature of major depressive disorder. Emil Uffelman (Complex Trait Genetics, VU Amsterdam), ended the symposium with a talk on estimating disorder probabilities based on polygenic scores using the Bayesian polygenic score probability (BPC approach).
In the second part of the morning, we had our second symposium Cardiovascular and Neurodegenerative Disorders. The symposium started with a talk by Rada Veeneman (department of Psychiatry, AUMC, location AMC) on genetic strategies to elucidate the complex relation between mental illness and cardiovascular disease. The second presentation by Alex Salazar (Department of Human Genetics, AUMC) focused on using long-read sequencing to identify structural variants associated with neurodegenerative diseases. The last talk in this symposium was one by Roddy Walsh (Experimental Cardiology, AUMC), and was titled: “the complex spectrum of dominance and recessiveness in genetic cardiomyopathies: insights from research in understudied groups”. After hearing, and getting inspired by, these interesting talks, we continued with discussion groups led by Dennis van ‘t Ent. We discussed our goals for outreach and collaboration within the network itself but also with outside organisations such as ABC, a summary of which will be distributed later. These useful discussions will be used to shape the network further in the year that is to come.
After the break, we continued with an inspiring talk from our keynote speaker Peter Visscher on exploiting within-family segregation variance to study complex traits. The keynote started with a brief recap of “old fashioned genetics” such as linkage, followed by novel work incorporating polygenic variation. The keynote lecture was followed by two sessions of 5-minute flash talks by 13 junior career researchers that introduced ongoing or very recent work. We ended the day with a symposium on the future of genetics, with Sjoerd van Alten and Tinca Polderman. Sjoerd van Alten talked about selection bias in genetic datasets and GWAS, and his solution of inverse probability weighting in UKB. Lastly, Tinca Polderman talked about involving people with lived experience in genetic research, illustrated by her work in autism research.
We thank all attendees for joining us on this day, and hope the full program will lead to novel collaborations within the network. A big thanks to Anne Landvreugd, Nikki Huberts, Tanya Phung, Rada Veeneman, Jorien Treur, and Margot van de Weijer for organizing this day, and to Amsterdam Brain and Cognition for funding the day.