Professor Zosia Miedzybrodzka: Expanding genetic testing for rare diseases
Image Courtesy of Zofia Miedzybrodzka
"Click here to listen to the full podcast episode"
In her time she’s seen genetic testing for rare diseases undergo huge changes, from looking at a picture of whole chromosomes taken down a microscope, known as a karyotype, to figure out if anything looks strange, through to testing single or a handful of genes, and then using what’s known as SNP microarrays to look at snapshots across the whole genome.
Today, technology has advanced and costs have reduced to a point where genetic testing services can use DNA sequencing techniques to read all the regions of the genome that encode the recipes for making proteins, known as exons, with a technique called exome sequencing. But, as I discovered when we started chatting, that doesn’t necessarily give you an answer straight away.
Zosia: The challenge then becomes to tease out the hundreds of thousands of variants in a normal person's genome and work out which single one might be the cause of the problem.
Zosia: So the way we do that, and this is the approach that's been developed with a major contribution for the MRC unit in Edinburgh and Dave FitzPatrick's team, was you sequence all of the exons of the DNA. And that represents about 2% of the entire genome. And compare it with what's the standard sequence for a person, and then come up with a file of variants. And then rather than look at all the genes in all the exons, filter that down to the genes that are known to cause the disease you're interested in.
Zosia: So the one we've done most of is intellectual disability. So you would focus down on it, and there's now a really well sorted out list of 5,000 or more genes. And we look for differences in those genes that are all known to cause learning disability, and then select out which variations are strong enough of a difference to be disease-causing.
Zosia: And then the clinicians and the scientists have to look at the patient together and decide if that difference actually looks like the patient in front of them, and that's the final test of whether that is the cause or not. Plus whether that difference has been seen before in other patients with something similar.
Zosia: And a key thing is, instead of just testing the patient in front of us for these complex tests, we try to do tests with three individuals from the family. Typically, it might be a child with their parents and where you get a new gene change that's present in the child, that's not present in either parent, that makes them look like other children with similar features, then you're kind of onto gold. And that's what we're looking for. And those are what we call de novo changes, and they're far more common than we ever thought they were. And it's the power of this technology that's letting us find that.
Kat: It's fascinating to hear this because, I know that there are a lot of variants in the human genome - there's no such thing as a normal genome. And the complexity of trying to look at this data and figure out what is actually not right compared to what is just a regular variation. Where is this technology going now?
Zosia: So going forwards, what we've been testing out is the idea of sequencing whole genome. And in Scotland, we've done that in a couple of ways. We've used a system called Short Read Technology and we've done a lot of testing with that. We've sequenced one and a half thousand Scottish patients' genomes as part of the 100,000 Genomes Project, and then used the 100,000 Genomes Project's interpretive systems to create a variant call file, the differences in the pattern, and filter down to the genes of interest.
Zosia: What I want to emphasise is that these advances that we've made in Scotland have been possible because of unique collaborations between the universities: Edinburgh, of Aberdeen where I work, of Glasgow, of Dundee, and NHS teams across the country who've worked really hard to take these systems, to put them into routine care and deliver the answers for patients.
Zosia: What I want to do is thank not just all the patients that took part in the study, but all the scientists and all the clinicians who did extra work above and beyond their normal job, often using systems that were just in the process of development, to test out these systems and come up with what is now a world leading service.
Kat: This sounds absolutely fantastic, but you know and I know that this stuff costs money, and it takes time. Is it actually worth it to take this approach to analysing these patient?
Zosia: So the first thing to do is pick the patients it makes most difference for. So in our clinic, for example, a common test we do is for familial breast and ovarian cancer. We can do what we call a panel test that looks at a relevant half a dozen or so genes four, about 200 pounds. If on the other hand, there's 5,000 possibilities like there is with intellectual disability, you're better to do that whole exon test that you can then interpret. But even go back and look at the sequence again in a few months time as more genes are being developed and find more.
Zosia: So what we've done is we've compared the performance of the whole exome system with the whole genome system, compared to our standard historic testing. And what we found is that cause of the clinical costs of having to keep looking and because most patients end up getting the Trio exome test done anyway, it's cheaper and much quicker for the patient to go straight to do the Trio exome test, rather than do two or three tests on the way.
Zosia: But as we go forward, if the cost of genome analysis falls, if the systems for handling the data become less clunky, if they show up less false-positives to the lab staff and the clinicians, it may be that one will overtake the other.
Kat: That’s Professor Zosia Miedzybrodzka from the University of Aberdeen.
