Dr Carl Anderson, Wellcome Trust Sanger Institute
Dr. Anderson:: So, I don’t really like the term “gene for” because we all have that gene and that gene exists in our genome to encode a protein which has a certain function and the function of that protein is critical or is useful to everyday human life. That’s why the gene is there, that’s why the protein is made and that’s why the gene exists.
What happens in certain individuals who have particular genetic conditions, and where this term “gene for” comes from, is that there are specific mutations which occur in that gene and it’s those mutations in that gene that change the function of the gene. They change the way that protein functions, the abundancy with the proteins present in our system, the folding of the protein that makes some subtle change to that protein that then changes our risk.
So it’s not true that this is a gene “for” breast cancer, let’s say. That’s not why that gene exists. It’s not present in our genome to cause breast cancer. It has some other function that when there's a particular mutation in that gene that then increases risks for breast cancer, whatever the disease happens to be.
Kat:: So, we should start talking about a “mutation for” breast cancer or heart disease rather than a “gene for”?
Dr. Anderson:: Exactly because we all have that gene and quite likely, if we didn’t have that gene, you’d be in quite a lot of trouble.
Kat:: And what sort of diseases would you class as a single gene disease and what sort of diseases are these multi-factorial diseases?
Dr. Anderson:: So the single gene diseases are things like cystic fibrosis, Huntington’s disease. They’re the type of diseases that actually, when you study genetics at school then these are the type of diseases, which you get taught. So, as genetic diseases, they are the ones where you have the pedigree diagrams and you look to see whether mum has the disease or dad has the disease and how many of the children have the disease and you could see a very clear, inheritance pattern. These are the classic Mendelian diseases, single gene diseases.
The more complex diseases are probably the once that have a greater frequency in the population so things like autoimmune diseases, type 1 diabetes, Crohn’s disease, Coeliac disease and things like most cancers, things like hypertension. And not just diseases you know, there are genetic influences on lots of human traits so your height, your weight, your overall body shape, all these human traits have genetic influences.
And assessing how important genetics is to each of these traits is a complex thing to do and as one of the jobs of the human geneticist. And once we’ve found out that genetics has a large role to play in some of these diseases, our next job is to go on and identify those specific regions of the genome that look to be underpinning that genetic risk.
Kat:: So how on earth do you do that? How do you pin down region of the genome or a particular gene and say that’s involved in heart disease, that’s involved in schizophrenia?
Dr. Anderson:: It’s surprisingly simple actually so basically, what we do, is we take a whole bunch of people who have the disease and a whole bunch of people who don’t have the disease. So how big is a bunch? Well, it really depends on the size of the genetic effect that you’re trying to find. So if you think, the particular disease you’re working on is likely to have genetic effects which are a relatively small effect, let’s say, then you’re going to need lots of people to do this experiment in.
So in the last five years or so, we started doing this genome wide association studies and so, what typically happens in a genome wide association study is we take say, 3,000 people who have the disease and we compare them to 3,000 people who don’t have the disease. And we genotype them and a whole bunch of just various positions throughout the genome. And so we end up having, say, half a million or a million sites throughout each of these people’s genomes and then we just look to see if any of these particular sites, there’s a difference between the people who do have the disease and the people who don’t have the disease.
And then if we find a site where there is difference, which look in the reference databases that we have to find whereabouts in the genome that you know, that particular site is, what genes lie in the area and does that actually tells them something about that particular disease.
Kat:: Presumably, you’re not looking at these millions of different variations by eye with pencil.
Dr. Anderson:: No.
Kat:: How do you analyse this kind of data?
Dr. Anderson:: So, the type of science that we do at Sanger and these genome wide association is extremely high–throughput, in the sense you’ve got lots of genetic data across many thousands of individuals. So the data sets are very large so we have to use computational tools basically to go in and conduct statistical tests at each one of these sites. And then when we’re analysing the results of that, we have to bear in mind how many tests we’ve performed to try and make sure that obviously, you do many tests so the chance of you finding a false positive is quite high. So you need to control for all that so basically, we use the computers to allow us to analyse all these data very quickly and efficiently. And we use statistical methods to try and pull out from all those many thousands and thousands of tests the few interesting sites that remain.
Kat:: Where do you think this kind of research is going to be heading in the future?
Dr. Anderson:: We’ve been relatively limited to the amount of a particular person’s genome that we can survey. So before, we’ve been just genotyping specific sites in the genome perhaps a million sites and using those one million sites throughout the genome to try and infer what’s going on throughout the whole genome. Now, with the advent of next generation sequencing, we can get hold of virtually every single base in any particular person’s genome. So we’ve got much more thorough coverage of one person’s genome. And so, this really increases the power of our studies.
Also, with the falling costs of those technologies, we can start to do that very thorough survey on many thousands and thousands of people. And so I think, what this is going to allow us to do is it will basically allow us to survey more of the genetic architecture of diseases. Before, we were limited to genetic variants, which were perhaps quite common in the population whereas now, we can actually start to survey down to genetic variants which are perhaps very rare in the population. And indeed, we can probably even get those specific variants which are unique to any one individual person.
Kat:: Do you think that one day in the not too distant future, when the baby is born it will have some blood taken and it will have its genome sequenced?
Dr. Anderson:: I think that’s very extremely likely, I think the potential medical benefits are quite huge, and I really think that we’ll be there soon.
Kat:: That was Carl Anderson from the Wellcome Trust Sanger Centre.