Peter Donnelly

I had the good fortune to attend the amazing Personalized Medicine World Congress put on by Tal and Gadi Behar at the Computer History Museum in Mountain View, California on January 26-28 of 2015. Part of my good fortune included a chance to do a video interview Peter Donnelly, the Director of the Wellcome Trust Center for Human Genetics at the University of Oxford. We chatted about genomics and personalized medicine, of course.

Here is a link to the video.

And, here is the transcript of our conversation:

PS: I heard one of the other speakers at this conference say that we have pretty much nailed the analysis of the genome and that really isn’t the challenge. The challenge is the phenome or the expression of the genome. First of all, I want to know if you agree with that and second, talk to us about the challenges of understanding what the gene sequences actually mean to me as a clinician and me a patient.

PD: So, I would largely agree with that. We have gotten much, much better at making sense of data that comes off of genome sequencing machines. Those machines do complicated experiments and produce data from which we can learn about or read a person’s genome. And, we are pretty good at that. That is not a completely nailed down problem, but we are reasonably good at it. The real challenge, as you say, is to take that information and work out what it means for individuals in terms of their risk of disease and what it means for patients and their doctors when they are sick and helping them in treatment. That’s a major challenge for us and I think we are only at the tip of the iceberg in terms of what is possible and what I hope will be possible in clinical medicine over the next five, ten or fifteen years.


PS: What are the kinds of approaches are going to be used? I have heard people talk about whether we will be able to get some understanding of that through big data and big data analytics. There was another presentation here [at the conference] about having “omics” profiles, where you would be able to have a profile of not just of your genomics, but also your proteomics and all the other “omics” that we have out there. Tell us how you are approaching it.

PD: I think, ultimately, it is going to be an empirical challenge for us. We are moving to a world where, at the moment, relatively small numbers, but soon larger and larger numbers, of people will have their genomes studied and often sequenced for research or as a part of their clinical healthcare. So within maybe five or ten years, there will be an increasingly large number of databases which link genomic information—all the information in someone’s genome—with other information about them, maybe medical tests or, in some cases, their whole medical records. So, I think we have to think about a world where that will be the norm. There will be tens, hundreds of thousands, or even millions of people with genomic data and healthcare data in those large databases. And the challenge and opportunity for us are to extract the information from those databases to learn how particular changes in someone’s DNA might make them more likely to react well to this drug or suffer if they have that kind of treatment rather than another. I think some of the ideas of big data and big data analytics will be important, but genetic data has a particular kind of structure and so some of the expertise we are developing in analyzing genetic data specifically will be important.


PS: How does the environment and other things that we think of as external to us as human beings play a role in the expression of the genome?

PD: It is really clear that many things about us in health and in our illnesses, our diseases, are a combination of the genetic predispositions we inherit and external factors like diet, lifestyle, and the environment. We don’t understand that very well at the moment. We know they are both important. Genetics is part of the story, but not all of the story. We know for some conditions, there are clear environmental risk factors: smoking and lung cancer, exercise, and heart disease, and so on. But actually unpeeling those and in particular working out the interplay between the environmental factors and genetic risk factors is something that we are just in the early stages of. As we amass more data and we get better data on individuals, on their genomic information, on the sorts of diets they have, exercise, and so forth, we will be able to untangle that better.


PS: I will close with a question that came up when I was in medical school with respect to Huntington’s disease. We used to try not to analyze someone’s risk for having it, although they have a family member with the disease because it was a devastating disease and there wasn’t anything we could do about it. This certainly is going to come up in the context of these genetic evaluations. Are people in your center thinking about those kinds of issues with respect to analyzing the genome?

PD: Yes, as you say, there are very real issues for people in terms of what information they would want to know from their genome and what they wouldn’t. Different people have different views on it. In some cases, that information will be helpful for them medically and in others, it won’t. In the same way, people have different views about other medical tests, whether to have them and how to get the information fed back. They as individuals, their clinicians, and us as a society need to think hard about those kind of issues. There is not going to be a one size fits all answer. It is a question of balancing the potential benefits of knowing the genomic information with some of the risks and the stresses and the anxieties that it might cause.


PS: This has been an interesting conversation. I will give you the last word. What is your prediction for frontline clinicians about how they will be using genetic data going forward? Is this going to be built into our electronic health records and pop up when we see a patient or is there something else that you think will happen?

PD: I think there will be differences in different areas of medicine and different medical specialties. In infectious disease, where the bugs that make us sick have genomes as well and also in cancer where cancer has a genome different from the genome the person has, these are two areas where in the next few years I think genomics will already be starting to impact the practice of medicine—it is in some of those cases already. Over the longer term, in many other areas of medicine, the genomic information will one of the tools, a part of the information that clinicians use. Obviously, a challenge for us is first of all to help us and the clinicians understand that information and then provide it to them in ways that they can factor it into their choices about treatment and diagnosis.

You can find all of the videos from the Personalized Medicine World Conference and other conferences we have filmed on our YouTube Channel:



All comments are moderated. Please allow at least 1-2 days for it to display.