Jamie Ellingford Interview Two

Hi, I’m Fran Hooley, Deputy Programme Director of the Clinical Bioinformatics PG Cert here at The University of Manchester, and I’m here with Jamie Ellingford today, just to chat through what’s happening with clinical bioinformatics currently. So Jamie, would you like to introduce yourself? Sure, so I’m Dr. Jamie Ellingford. I’m a Research Fellow at The University of Manchester and I’m based at the Manchester Centre for Genomic Medicine. So we last spoke when you were doing your PhD. So what have you been working on in that time since then? So that seems a long time ago, but I’ve stayed in the same seat.

I’ve saved working with the same people but actually the discipline and the types of things that we do have changed drastically over the past five to ten years. And whilst there’s still an underpinning emphasis to identify genetic variants that are causes of particular disorders, lots of the techniques that we use and our capability to do that has evolved greatly since we last spoke. So in your opinion, you know, what has changed in clinical bioinformatics over the last five years?

So one key thing, which has become more and more available as part of genetic or genomic medicine research is the availability of big data, and so the fact that we now have access to a huge number of genomes for individuals with particular rare disorders. We’ve also expanded the search base that is used as frontline diagnostic tools and research tools for people with these genetic conditions.

And so traditionally we’ve used more targeted sequencing based approaches, we’ve looked at specific regions of the genome, but now for lots of disorders the frontline test is to sequence the whole entirety of the human genome and to identify as many variants as we can across the complete spectrum of the 3.5 billion nucleotides that are contained within each of our genomes.

Great stuff and I know we’ve had challenges without, you know, continue to have challenges about data storage and interpretation. What do you think’s happening now? How are the data storage issues being addressed currently? So data storage as we’ve just alluded to, we have much bigger data sets both in terms of number and in terms of size of those individual data sets. And so what’s quite apparent is that storing this data in single institutions is quite challenging but developing ways to share data, to work within cloud-based systems has become a hot topic over the last few years.

And whilst we still haven’t found the perfect way to do this, those sorts of things are in the conversation as to how we tackle these challenges going forward.

So in terms of interpretation, have there been any major changes over the last five years? So this shift from targeted tests looking at specific parts of the genome to generating whole genomes for individuals with rare disorders really has required a paradigm shift in the informatic approaches, but also some of the follow-up experiments that we do on the back of this data. We now have the opportunity to identify the complete spectrum of genetic variants that cause genetic disorders, but our knowledge and our understanding of how particular parts of the genome work and how they can be disrupted and cause genetic diseases isn’t quite there yet.

And so, whilst we’ve started to develop informatic ways to computational processes to analyse this whole genome data in a better way, those aren’t fully established and they’re not so mature. That being said, since we’ve spoken there’s actually been the release of formal guidelines to start to bring together the community as a whole and worldwide to a standardised approach as to how we actually interpret genetic variation and to ensure that there’s an evidence base for all of the decisions that are given out of clinical diagnostic laboratories. And do you think there are variants missing because we don’t know how to interpret them? Absolutely.

So those guidelines that I’ve just alluded to were built specifically to interpret variants which affect the protein coding regions of genes, so those parts of genes which are turned directly into protein. But now with whole genome data sets, we can look at non-coding parts of those genes that affect the way that they’re switched on or affect the way that these special molecules called mature messenger RNA are created through a process called splicing. And our understanding of how variation can disrupt this gene expression and splicing process isn’t quite as mature as that that we know for the protein coding parts of genes.

And so these are the types of variants that we do miss but we have the opportunity to start to build both computational programs to better predict which of these variants are disruptive but also extra techniques within the clinical laboratories to prove that these these variants are having the effect that we’re predicting them to do. Great stuff. So we’re seeing more and more mass sequencing projects from the hundred thousand genomes projects through to the NHS 5 million whole-genome plan. Can you describe your work that’s been involved in the hundred thousand genomes Project?

I was fortunate enough to be involved in the pilot project of the hundred thousand genomes project and we published these findings in November 2021 as part of a large National Consortium.

Amongst other things, one of the key findings from this study was that for 25 percent of the individuals that we recruited, we were able to find a genetic diagnosis, although that did vary depending on the specific disorder that the individuals were recruited with. Importantly because this project was so aligned with routine clinical care and NHS clinical Diagnostics. We were also able to show the utility of integrating genome sequencing into the clinical care pathway for these individuals. And there’s some really striking examples of where there was real benefit to being able to analyse a genome and to find the genetic diagnosis for these individuals.

And to talk about one of those examples from my own research, in a young family, where an individual, a young boy was first recruited to test him with severe visual impairment. We were able to find a genetic diagnosis that gave key information as to how that disease would develop. Importantly, it suggested that his kidneys could start to perform abnormally in his late teenage years and at the time he was 14 years old and he was referred to a renal specialist here in Manchester.

And as a result of being able to pick up those signs early, it meant that the team were able to manage his care both by putting him on hemodialysis in a timely fashion and also being able to perform transplants before end-stage renal disease, which is a really poor indicator of outcome. And so in this situation that genetic diagnosis was really able to transform the care that this family received and to best manage them in the clinic. Fantastic. So you’ve done some work with nanopore sequencing. How do you see this impacting labs? So a lot of the things we’ve already talked about have been through the generation of data using short read technologies.

Nanopore offers us the opportunity for long read sequencing and actually I think provides advantages or potential advantages in lots of different areas. So, a broad advantage of this is that it could increase the utility and the spectrum of things that can be tested for in clinical diagnostics laboratories. Although we’re not quite there yet, some of the areas that I think this could be a impactful in include the ability to pick up things like methylation status, both for DNA and RNA molecules, the ability… the greater ability to identify copy number variants and structural variants and a new ability to identify really complex structural variants, which are really difficult to piece together from short read technology by itself.

And another area where it may be impactful includes the ability to characterise whole transcripts in single reads, and so being able to characterise full-length transcript isoforms both gives us advantages in determining what types of genes and isoforms are switched on in a cell but also to quantify them with greater resolution. And so I think all of these techniques whilst not necessarily something that we talk about every day in a clinical diagnostic laboratory at the moment, maybe in the future Nanopore could provide lots of advantages in these areas.

So looking at the future then, as genomics moves into the realm of big data do we need new ways of storing that data to enable us to ask those big questions and how important will clinical bioinformaticians be within that? So we certainly need solutions to be able to both store and access this data efficiently and clinical bioinformaticians I think hold the key both from an informatics side, but also from a biological interpretation side. We’re now dealing with data that the NHS just isn’t used to looking at every single day and the clinical scientists will both on the bioinformatics side and the genomic side will need to develop those informatic capabilities to be able to analyse these data sets most efficiently.

So with the covid pandemic we’ve seen more mainstream media reporting on sequencing and variance. Do you have any thoughts around the need for more public engagement to explain what these terms actually mean? I think I said this to you when we spoke during my PhD, but public engagement is absolutely critical for what we do, not only the general public but also patients, as well as key stakeholders that are involved in the research that we do in Manchester, both by funding it, by participating in it and by providing lay opinions of the work that we’re doing. Their input is absolutely critical for what we do and we as researchers need to ensure that that social responsibility aspect.

of our work is still at the forefront. Of course, in a modern-day, there’s lots of opportunities to do that, and as people have become more accustomed to talking online through the pandemic it makes these things I think, in some ways a lot easier. Recently for example, I’ve done an Instagram take-over for a charity who fund our research and coupled that up with lots of content on Twitter to explain our work in simple terms as to what we’re trying to do and the impact that we think that this will have both on patients and on society.

That’s great for us. It shows, from a selfish point of view, it often shows that people are interested in your work and would like to find out more about it. It also gives us an insight into what the general public want from us as researchers. And I think that in terms of planning and how we go forward is really critical for our thinking as researchers. Fantastic. Well, thanks Jamie. It’s been great to catch up. And yeah, thanks again for your time. No problem.