Novel Sequencing Technologies in NGS for Genomics Research

0:06 
Thank you everyone for joining today's Monthly Science Exchange, which is part of our Omics series. 

 
0:13 
Thank you for joining. 

 
0:14 
I'd like to introduce you to today's discussion group leaders. 

 
0:17 
So all right. 

 
0:21 
Yes, today's discussion group leaders are Victor Turcanu, who is a Senior Lecturer in Allergy at King's College London, and also Jiannis Ragoussis, who is the Head of Genome Sciences and Professor at the Department of Human Genetics at the McGill Genome Centre. 

 
0:35 
So, in regard to how the session format will work is Victor and Jiannis will both give a short presentation each, of around 10 minutes. 

 
0:44 
And then we'll open the floor to an open roundtable discussion format. 

 
0:51 
Currently you can only use the chat, you don't actually have audio and video permissions. 

 
0:55 
So what will happen is once both presentations have concluded, I'll then quickly go to the attendee list and give you a request to join as a panellist. 

 
1:03 
So please note, this will be enable you to join with audio and video. 

 
1:06 
If you do not want to join with this, that's no problem at all. 

 
1:08 
If you just want to listen, that's fine, but what we'll do is we'll give you permissions to do so just in case that is of interest to you both or to everyone. 

 
1:18 
So thank you very much. 

 
1:20 
And I'll now hand over to Victor and Jiannis to begin the session. 

 
1:23 
So thank you, over to you guys. 

 
1:26 
Thank you. 

 
1:27 
Thanks for your introduction, Declan. 

 
1:29 
So brief introduction as I'm starting to share my screen. 

 
1:35 
I am a clinician scientist, I trained in internal medicine, then I did a PhD in Immunology. 

 
1:42 
So I've been doing allergy immunology randomised clinical trials for the past 20 years at King's College in London. 

 
1:51 
And I'm going to briefly give some examples based on my work about using next generation sequencing to explore Peanut Allergy Immunotherapy. 

 
2:03 
So this is very translational. 

 
2:06 
That's what I'm doing. 

 
2:07 
And as I was telling Ioannis and Declan earlier, I am excited by the new opportunities that are opened by this new technology. 

 
2:22 
So let me go to the next slide to show you briefly for those who are not aware that peanut allergy is the leading cause of death by anaphylactic shock. 

 
2:42 
It is not the most frequent allergy, but it is causing significant difficulties because it lasts for the entire life, unlike milk and egg allergies that normally lost spontaneously 95% of the cases, peanut allergy remains as a disease and rarely outgrown, only about 20%. 

 
3:12 
Most importantly, its prevalence has increased in the UK and other developed countries during the 1990s in the frame of the allergy epidemic that affects all these countries. 

 
3:23 
So it is quite significant, and we see such tragic news in the newspapers again and again about teenagers often going to university, eating with friends in a takeaway and unfortunately died of an anaphylactic shock because of their peanut allergy. 

 
3:51 
So, one of the great successes in peanut allergy resulted from work in our group at King's College London, we introduced peanuts early in life in a large number of children in a randomised control trial because we wanted to test the hypothesis that we can prevent peanut allergy by early life oral polarisation. 

 
4:17 
So this is mainly the paediatric group at King's College London with other colleagues from different organisations. 

 
4:28 
And we carried out this study that was reported in 2015 in the New England Journal of Medicine. 

 
4:37 
What we did is that we recruited 640 infants between 6 and 10 months of age who had severe eczema or egg allergy, because statistically that means their risk of peanut allergy was about 10 times higher than the general population. 

 
4:55 
And that allowed us to keep the size of the study small. 

 
5:00 
We had that randomised into an intervention group who are eating peanuts products, not peanuts because they're dangerous obviously to give to infants, but peanut products early in life like in countries where peanuts are consumed early and another the control group, babies who would be avoiding peanuts. At the time 

 
5:24 
the advice was to avoid peanuts until three years of age and at five years of age when most people would have been emerged, we noticed an 81% relative reduction with 17.2% babies becoming peanut allergic in the avoidance group but only 3.2% in the consumption group. 

 
5:48 
This led to a complete reversal of the advice. 

 
5:52 
So babies were advised well, parents were advised to introduce peanut products into baby’s diet as soon as reasonable after the end of exclusive breastfeeding, which would be roughly after six months of age or depending on how long exclusive breastfeeding is done. 

 
6:15 
And it is also demonstrated that it is safe, it is effective in all races as far as we could see it in London. 

 
6:25 
And the immunological basis of prevention is oral tolerance induction. 

 
6:29 
However, there's still the 20% of individuals who probably would not be able to prevent their allergy as well as those who are already peanut allergic. 

 
6:40 
So what can we do for them? 

 
6:43 
And it is immunotherapy for peanut allergy that is our next focus of interest. 

 
6:49 
And I have to stress that peanut allergy is driven by T-cells and this is why T-cells are the focus and next generation sequencing, especially looking at single cells is such a powerful and useful tool. 

 
7:05 
So in mice, regulatory T-cells from tolerant mice can treat the food allergy, in this case ovalbumin, the main allergen from eggs. 

 
7:17 
So if a mouse is rendered tolerant to eggs by oral exposure by eating one milligram of ovalbumin per day for five days, that mice will develop regulatory T-cells in the mesenteric lymph node. 

 
7:33 
It's a drain the gut and if these cells are transferred to an egg allergic mouse, then that mouse will become tolerant to eggs. 

 
7:44 
This is true not only in mice. 

 
7:46 
Obviously you can do whatever you can design with mice, but even in people - you can cure peanut allergy by bone marrow transplantation. 

 
7:55 
So in this case, a 12 year old boy with severe immunodeficiency had peanut allergy and an anaphylaxis reaction and this boy had to be treated with bone marrow transplantation for the 

 
8:12 
immunodeficiency of course and that led to the resolution of peanut allergy and a negative peanut challenge. 

 
8:20 
Naturally bone marrow transplantation has a high risk of death. 

 
8:26 
So obviously one would not recommend for peanut allergy, but this demonstrates that simply by transferring new set of T-cells, regulatory T-cells that can be sorted. 

 
8:41 
So what does next generation sequencing and what does single cell strategy add to us? 

 
8:50 
Well, we can realise that we have multiple subsets of Tregs in from peanut allergic in red that are clustered distinctly from the Tregs from the healthy controls. 

 
9:05 
And at the same time we're looking at the response between in our CD154+, that is a marker of activation and control cells that are demonstrated as being activated at the same time. 

 
9:24 
So we identify cells, we can identify the cells that cause allergy, we can identify cells that stop the allergy and hopefully we can design immunotherapy protocols that would enhance the production of these T-cells. 

 
9:47 
And so basically, we designed an experiment looking at children who underwent oral immunotherapy. 

 
9:57 
And we used the next generation sequencing on single cells, looking at single cells collected at two different type points, box cells as controlled non-specific cells so that we could compare them. 

 
10:12 
And we took before the start of immunotherapy and 18months into the therapy in order to identify potential clustering as we saw. 

 
10:25 
So the signal was quite good. 

 
10:27 
We had a fair number of reads, a good percentage of total alignment and a reasonable read depth with the downside, however, most of the cells seem to cluster in terms of different time points. 

 
10:49 
So these are disorder and time point and the other time point the immunotherapy and after one year. 

 
11:11 
Actually that was not sufficient for us to define... 

 
11:18 
But in the literature that must become possible and I'm going to quote where from other allergists work in this area. 

 
11:29 
So this is the group of Canada in the United States different, you know therapy for peanut allergy. 

 
11:40 
This group managed to separate enough specific cells into single wells and analyse single cell in types of Fluidigm and that limits the number of genes. 

 
11:53 
But it is a very convincing picture with the clusters emerging in terms of the cells that respond. 

 
12:03 
So it is an interesting finding, and it gives some insights that we could not have previously. 

 
12:11 
It is not only one cluster of cells, the regulatory T-cells and one cluster of the allergy driving T-cells, but there are actually 7 clusters. 

 
12:20 
And as we see from the beginning of your 30, cluster 7 almost disappears whereas cluster 6 expands almost so much as to dominate the risk. 

 
12:32 
And this is the power, this is a translational element of single cell next generation sequencing that we can monitor in real time what is going on. 

 
12:44 
The opposite is also true. 

 
12:46 
This is where from Eric Wambre also in the United States, looking at who are the culprits for driving peanut allergy and other allergies too. 

 
12:58 
And it is very clear example of identifying the Th2A subset that causes it and another subset, the T follicular helper 13 substrate identified by a second, a separate group. 

 
13:16 
And apparently these are the culprits driving anaphylaxis and potential death in allergies. Peanut responding T-cells of course display multiple different transcript terms. 

 
13:29 
And by putting the key chains differential express genes, one can get an idea of the types of cells that intervene. 

 
13:39 
And that's again a different group reporting on this. 

 
13:44 
But I would like to finish here and point out that all this work has been done in the group led by Professor Lack and also Professor Santos, Professor Brough and Doctor Chan, as well as colleagues from different departments, including the Genome Centre at Guy’s and St Thomas'. 

 
14:05 
So thank you and I will stop sharing and I'll let Jiannis share some of his work in this area. 

 
14:19 
Thank you, Victor. 

 
14:33 
OK, I hope you can see my screen. 

 
14:36 
So thank you for the introduction, Declan, and thank you, Victor. 

 
14:44 
I'm Jiannis Ragoussis, Professor in Human Genetics at the Department of Human Genetics at the Department of Bioengineering at McGill University and Head of Genome Sciences at the McGill Genome Centre. 

 
14:58 
And I'm also Associate Director of Canada's Genome Enterprise, which is Canada's main sequencing and bioinformatics analysis alliance. 

 
15:10 
So what I'd like to go through a little bit in terms of, you know, stimulating further discussions is the following that you know in the last couple of years, many different new sequencing technologies have been introduced. 

 
15:34 
So you know in a sense, at last you can say, there are many different competitors in the next generation sequencing market offering different instruments. 

 
15:47 
So we are spoiled for choice at the moment, which can be good. 

 
15:55 
We have to, as shown here, you know, we have to navigate, quite crowded sequencing landscape. 

 
16:07 
So just to show you know, some of what has been introduced recently, apart from Illumina who was the predominant force in short sequence followed by proton, then Element last year starting from 2022 as benchtop instruments. 

 
16:40 
We're going to that in the moment, followed by Ultima by the end of 2023 in December. 

 
16:51 
And then another benchtop instrument also was introduced by PacBio. 

 
16:57 
Then we had the new Illumina X sequencer, competitors from China with a GeneMind 

 
17:05 
Genolab M was introduced as another benchtop instrument, and then just a month ago, Complete Genomics introduced a new DNBSEQ-G8. 

 
17:21 
For the first time I produce a shortage, which can lead long. 

 
17:34 
And then we had newcomers. 

 
17:37 
So in technologies, new instruments from PacBio with a REVIO high throughput sequencer, the Illumina complete long reads and then Nanopore with sort of small and agile high throughput sequencers like the P2 and v14 chemistries. 

 
18:01 
I'll go into that in a moment. 

 
18:04 
So memorise what how the landscape looks like or just a snapshot what I have here on the left, there are three benchtop instruments which are very powerful. 

 
18:20 
So you can see here from Singular G4, the PacBio also and the Element AVITI. 

 
18:25 
And what is interesting here is that these instruments have also introduced a new and improved sort of either improved and new sequencing by synthesis approaches like singular with improved sequencing by synthesis element using avidites to perform sequencing by synthesis combined with running circle amplification and PacBio also introducing new approach which is sequencing by binding. 

 
19:06 
And all these aim at high accuracy in terms of the sequence that is being produced. 

 
19:15 
And they are medium throughput instruments. 

 
19:19 
They seem to be perfectly good for single cell work, for example. 

 
19:28 
On the other hand, they are you know, I put here on the right - two new instruments that are big ones. 

 
19:38 
So they are industrial scale if you want, from Ultima in particular. 

 
19:47 
And what is new about this technology is that the sequencing occurs through natural basis. 

 
19:57 
So it's sequencing by synthesis again, but mostly with natural basis spiked with fluorescent bases and there are no terminators. 

 
20:08 
What does this do - is to reduce the price significantly of the sequencing in combination with very cost effective wafer technology away from microfluidic flow cells that all the others are using. 

 
20:24 
And what is ground breaking with this introduction is mainly the possibility to move not just into the $100 US genome, in terms of reagent cost, but also beyond that significantly. 

 
20:48 
So the other main introduction was Illumina Novaseq X/X plus also with improved sensing by synthesis technology and very throughput moving towards you know, sequencing 64 whole genome sequence. 

 
21:11 
So this is just what, you know, we have now many different and we should not ignore the complete genomics offerings here as well. 

 
21:23 
So there are many contenders, you know, and I feel that for example, the three on the left would be very well suited into, you know, clinical diagnostic laboratories, for example, and being used in hospitals as well. 

 
21:43 
Some examples very briefly on, you know, how this type of technology, and I'll focus on long read technology to complement Victor, can be introduced towards, you know, many translational projects I have here. 

 
21:59 
I'm showing an example from my lab's work on sequencing the HLA region. 

 
22:06 
This in order to get high resolution, a little specific and habitat specific sequences, we use long range PCR and combine this with long read sequencing. 

 
22:22 
And the reasons for doing that is that it can be done with both Pacific Biosciences and nanopore technologies. 

 
22:37 
Here we are showing an example with nanopore technology due to the speed. 

 
22:42 
So basically this approach can be used to rapidly HLA type at very high resolution samples from disease donors so that they can be offered to the matching recipient for transplantation. 

 
23:01 
And as you can see here with the nanopore particularly, we have managed to demonstrate that within four to five hours from sample collection, one can have fully analysed data using - in this case an instrument and combine this with live base calling. 

 
23:24 
This is with the new R10.4.1 flow cells. 

 
23:30 
But the nevertheless, as we know, nanopore sequencing has issues with base calling. 

 
23:38 
So how do we get high quality data? And I'm showing you here, you know, we use many different base calling tools to increase the quality of the sequence that has been obtained. 

 
23:51 
And indeed we can reach with, you know, Q30 just over Q30. 

 
24:00 
So this would be more or less similar to what the quality will get with four short read sequencing showing that actually there is good potential with long read technologies, both the long read technologies available at the moment to reach high accuracy. 

 
24:19 
Obviously we've never at the cost of how many reads you get, but as I'm showing you here, it is important because to use this these tools and the technology because very rapidly with the improved base calling we can get- if you go from left where you have an ambiguous calling with standard collars, the high quality collars can actually resolve both alleles accurately. 

 
24:50 
So it can offer correct typing and that allows the technology to be considered for clinical use in the very near future. 

 
25:00 
And that's work done from between my lab and another lab. 

 
25:07 
So yeah, at the hospital, McGill Hospital. 

 
25:11 
The other thing I like to highlight is you know, how these new long read technologies can fit into the clinical diagnostics here on the left. 

 
25:19 
They are very powerful in terms of resolving repeat expansions. 

 
25:25 
You know an example shown here with Cerebellar Ataxia, and you can see that by applying long read sequencing one can accurately determine the length of beat expansions. 

 
25:38 
So they are very useful for this approach, but also to resolve standard repeats. 

 
25:45 
And what is important and very powerful, you know, a feature of the long read technologies is that not only can you identify whether you have a repeat and that's shown here. 

 
26:01 
So we have a wildtype repeat, but then we can also check from the raw data what is the methylation pattern. 

 
26:09 
So we see the repeat, we can see whether it's expanded or not, but we can also check whether, you know, the surrounding region by sequencing fragments that are long - over several thousand base pairs - to identify directly from the signals the methylation patterns. 

 
26:30 
And you can see here why this is important. 

 
26:33 
We resolve the alleles with this technology and the wildtype is methylated only on the five prime part. 

 
26:42 
But you can see that the expanded allele on the right, which is the expanded allele, we look at the methylation pattern and it's completely methylated as you can see here. 

 
26:52 
So what is important is for also for many different applications, not only with long read technologies. 

 
27:00 
Not only can the particular mutation, if you want, be identified, but it can be assigned to a particular haplotype and it can also be assigned to particular methylation patterns. 

 
27:15 
And just very briefly, just very important publications that came out last year where these technologies due to the advantage of, you know, being able to rapidly produce sequence and mainly spear headed by the nanopore technology. 

 
27:30 
At this point, they are finding ways to combine next generation sequencing with either on the left, you know, the surgery room. 

 
27:41 
So in the operation room and you can identify and characterise samples derive from the brain tumours for example, and immediately inform within, you know, one hour and a half to immediately combine with AI applications to immediately inform what surgery should be taken, what type of surgery should be taking place in the operating room. 

 
28:06 
But also in the emergency room where within a couple of you know, one day or even shorter. 

 
28:13 
Now you can produce data identifying mutations that guide the treatment of persons in the operating room. 

 
28:25 
As so this the site on the right is the Stanford study and on the left in Holland by the reader. 

 
28:34 
And I'll stop here in the interest of time. 

 
28:38 
Thank you, brilliant. 

 
28:43 
Thank you very much Jiannis and Victor for your presentations. 

 
28:50 
So thank you both for your presentations. 

 
28:52 
So what we'll do is now we'll open the floor and begin the round table open format discussion. 

 
28:59 
So what I'll do is I'll quickly go for the attendee lists and you should see a pop up offering to join as a panellist. 

 
29:06 
Only accept this if you'd like to join with video and audio. 

 
29:09 
Obviously feel free to ignore this if you would not like to choose to do so and just want to listen instead. 

 
29:13 
But we'll just do one sweep of this. 

 
29:15 
And obviously later on in the session, if you do get the urge, just type in the chat. 

 
29:19 
And then what will happen is I will then be able to do it that way as well. 

 
29:23 
So I'll just go through the list and we'll see if there's anyone who does want to become a panellist. 

 
29:29 
So just give me a few minutes. 

 
29:30 
Yes once you accept this you can then just turn on your webcam and audio. 

 
29:38 
So also what you do as well. 

 
30:06 
If you do want to become an panellist, just type this into the chat and then we can just do it this way as well. 

 
30:12 
So there are quite a number of people in the chat. 

 
30:27 
I guess would you both like to begin to kick off the session anyway, perhaps maybe any questions for each other following your presentation? 

 
30:33 
Then what I'll do is I'll just give permissions to those who would like to join us. 

 
30:37 
Well, Mike's here as well. 

 
30:40 
So yeah, anyone who is also joining through video, feel free to introduce yourselves to Victor and Jiannis as well. 

 
30:44 
Be great to hear from you both. 

 
30:46 
Actually, we have Mike joined as well. 

 
30:49 
Yeah. 

 
30:49 
Michael Quail from Wellcome Sanger Institute. 

 
30:53 
Hello Victor. 

 
30:54 
Hello, Jiannis, Great talk. 

 
30:56 
Hi, Mike. 

 
31:11 
Yeah, I've done a sweep through the attendees again. 

 
31:13 
Maybe you guys perhaps kick off with a question or something like that. 

 
31:16 
And then what we'll see is, you know, people may be more eager to then sort of join in. 

 
31:20 
I think people are a bit shy, so perhaps yeah. 

 
31:25 
Do you both have any questions for each other perhaps? 

 
31:27 
Mike, do you have a question for the Jiannis and Victor regarding their presentations? 

 
31:33 
No, I'm fine. 

 
31:35 
I've seen Jiannis talk on many occasion, and you know, he's doing some great work, particularly with long read. 

 
31:42 
I guess one question I might have is, you know, how do the others see the balance between long and short? 

 
31:51 
You know, I think this debate has rolled for a while. 

 
31:53 
I think several people have said that long will dominate, but you know the date at which that seems to potentially happen seems ever further off with the development of newer and cheaper short read technologies. 

 
32:10 
How do you think that balance is going to play out and whether it's just going to roll into horses for courses? 

 
32:25 
OK, I'll start. 

 
32:26 
Thank you, Mike. 

 
32:28 
Great to see you. 

 
32:29 
So the, it's a very good question and that's something that, you know, we have to try and address. 

 
32:42 
The issue is that with the new sequencing technology, short read sequencing technologies that, you know, the cost has moved further away towards, you know, the, as we said, the $100 and less 

 
33:01 
genome while the longer technologies are far behind. 

 
33:05 
So they cost, you know, you could argue at least six times more. 

 
33:09 
So roughly speaking for a human genome equivalent. 

 
33:16 
So at least what we have observed and what we are doing is that the long read technologies are finding their way not only in terms of increased use for research purposes. 

 
33:36 
So they if you want, they get a bigger slice of the sequencing pie in terms of usage for many different projects and they are finding their way into the hospital. 

 
33:49 
So they are more and more actually clinical diagnostic labs that are working towards introducing, you know, long read technologies. 

 
34:00 
And I think it's interesting, it's going to be very interesting particularly if not only just nanopore, but maybe also Pacific Biosciences are going to deduce a real bench top type long read sequencer. 

 
34:15 
So if both, you know, main providers of long read sequencing are able to offer devices that are benched top, at least then that will enable the longer read sequencing technologies to move into, you know, many different uses in hospital settings in smaller clinical diagnostic laboratories. 

 
34:43 
So I think there are two things happening. 

 
34:45 
The part of the slice that is being, you know, taken by the long read technology is increasing, but the pie is also, you know, getting bigger so that more and more work is being done through next generation sequencing. 

 
35:01 
You know, there are so many assays now that even protein, you know, from a few years ago, they have protein, you know, based assays, high throughput Multiplex assays. 

 
35:12 
That Olink, for example, is a classical example, which has moved from, you know, microarray type output to next generation sequencing outputs. 

 
35:22 
So the sequencing, next generation sequencing is being used as an output for other assays as well. 

 
35:29 
So the whole pie is getting bigger and bigger, but I think the short read technologies are here to stay because of cost efficiencies. 

 
35:43 
And as Victor was saying, you know, single cell and spatial transcriptomic technologies are going to find their way into the cleaning. 

 
35:54 
And at this point, you know, it's fantastic to combine them with long reads, but it's not absolutely necessary in order to perform expression profiling. 

 
36:04 
So therefore, short read technologies will stay, I think as the predominant way to do things due to cost I think, right. 

 
36:18 
But Nanopore is, you know, as we know is finding it more and more ways into smaller labs and it has been really powerful in helping clinical diagnostic test development, and a range of new applications that other people haven't perhaps dreamed of before. 

 
36:41 
As you said, rapid clinical sequencing and mobile sequencing and all of the things that some of the other technologies perhaps don't appear to be able to do right out-of-the-box. 

 
36:53 
Yeah, it is amazing. 

 
36:55 
And the other thing with this technology is that, you know, you can bring them into the classroom even and introduce, you know, new generations of scientists into genomics and sequencing, which, you know, with the monsters we had previously wasn't straight forward to, you know, now you can sequencing in a few minutes if you want. 

 
37:20 
Yeah, I, if I may add here, I think that we need to add the third element here, which is the bioinformatics, because we still haven't solved the problem of IP low end. 

 
37:34 
We are sequencing huge amount of genes from a relatively small number of participants. 

 
37:40 
So the cost is important because is it better to sequence say 10,000 cells or look at 10,000 different samples with a shorter sequence or a smaller number for the same cost. 

 
37:59 
So again, it is a trade-off here and I think that everybody is aware of that. 

 
38:05 
But, I say that the hard problem of IP and low end is still one of a major barrier because if we start with the end purpose in mind, what we want to do is to distinguish between a biological characteristic of a clinical case or error research case. 

 
38:31 
And it is a trade-off between long sequencing and synchronous and shorter ones. 

 
38:39 
Yeah, I can add that for the bioinformatics which is one of the most frightening parts of introducing this technology into hospitals in particular. 

 
38:56 
What at least what we are doing is the following, to fully automate the analysis pipelines and incorporate, you know, all the recent fantastic developments, for example, in producing understandable reports using the reporters that have been published recently for the cancer genome for, you know, germline and somatic mutations for example. 

 
39:21 
And incorporate all this into a fully automated pipeline that generates reports that the clinicians can start looking at without, you know, panicking and then transplant these pipelines into the hospital basically. 

 
39:38 
So we, yes, it's very complicated, but then it's the job of, let's say the research scientist to simplify workflows and allow, you know, the clinical scientist to implement these workflows into their settings. 

 
40:04 
Hello, buddy, can you hear me? 

 
40:08 
Hello. 

 
40:09 
Yeah, yes, yeah. 

 
40:11 
Hi, this is Patrick Descombes from Nestle in Lausanne, Switzerland. 

 
40:14 
Hi everyone. 

 
40:15 
Mike, great to see you here. 

 
40:17 
I have two questions. 

 
40:18 
One is first of all, thank you for your presentations. 

 
40:22 
Jiannis, as you mentioned, you know the couple of new technologies which are appearing, I was wondering if any of you have hands on experience. 

 
40:30 
I know, we know that you know the AVITI system is being more and more placed to my knowledge very few ONSO, I don't know about Singular or Ultima, just if any of you has hands on in addition to the information we get either from publication or from the companies themselves. 

 
40:48 
And the second question I have is, there are situations where we want to sequence very few samples. 

 
40:58 
And of course, coming from the food industry, I'm not only thinking of Human Genetics, but also, you know, more food safety quality related topics, where having a fully automated system from sample to library ready that is extremely simple. 

 
41:14 
And we hear about all tracks of nanoport as an example of what would be a dream. 

 
41:20 
And I was wondering if any of the panellists has information to share on this type of devices? 

 
41:25 
Thank you. 

 
41:29 
Thanks, I can start then, I can tell you the following. 

 
41:36 
We have been working with PacBio on the ONSO instrument, to your first question and indeed what we see is that the data quality is incredible. 

 
41:57 
So basically the accuracy is in you know in the Q50 to 60. 

 
42:04 
So most rates are extremely accurate, and I didn't show this here. 

 
42:09 
I maybe you know, in the interest of time, I didn't put that in, but you can see how fantastic it is in terms of resolving also repeats in particular, and producing high accuracy, you know, data over repeat regions, trinuclear repeats and microsatellites in particular. 

 
42:35 
So there is a lot, you know, they do what they promise. 

 
42:41 
I think with the new chemistries. Thank you that's super interesting. 

 
42:50 
Thank you for sharing. And Ultima is really good. 

 
42:53 
I mean Mike has Ultima in his lab so he can talk about that as well. 

 
43:00 
I mean, yeah, your second question was for the food industry. 

 
43:14 
I'm not sure or any topic, sorry. 

 
43:16 
It's not good anything where we have 2-3 samples and we are in the field, you know, we talk about these portable devices 

 
43:23 
if you're outside. There are, there has been a lot obviously with nanopore technology which allows this to happen. 

 
43:35 
And there are, you know, the institute pastor has in other institutes had developed, you know, little suitcases where which you can take remote locations and using your laptop you can sequence in combination with a nanopore device. 

 
43:54 
So it is possible. 

 
43:55 
And I think what is great with particularly with a nanopore technology is you know that the adaptive sequencing. 

 
44:05 
So basically if you are targeting a particular pathogen, let's say in food that you want to identify, this can be done extremely effectively using adaptive sequencing approaches. 

 
44:20 
So Nanopore I think is ahead in that respect, sequencing a small number of samples. 

 
44:25 
And the other thing I would like to mention is that there are companies that develop automations to completely automate the process, in terms of using microfluidics like you know - for example Miraculous, there's the power blade we have been working on in Canada, which are very well suited in terms of processing small numbers of samples in a remote locations without the need of highly specialised personnel. 

 
45:02 
So I think the combination of new, you know, microfluidic sample prep devices with, you know, handheld sequencing devices is going to make a difference in this area. 

 
45:21 
Thank you very much. 

 
45:22 
Yeah, I think is kind of wider, doesn't it, because we've not only got, you know, we got a range of different sequences, whereas obviously on nanopore technology you've got things like Flongle. 

 
45:35 
So you can do a sub $100 sequencing environment if you only want a little bit of sequence on things like singular, you know, they have a range of different flow cells and multiple lanes to allow for flexibility of loading different samples. 

 
45:55 
And you know, if you look at the cost point of a single lane of one of those, I think that the cheapest is somewhere around the $200 mark. 

 
46:05 
So, you know, you could access a lane of, if all you want is a little bit of sequencing, you can access a lane of, and I think Element have also gone down that route recently, with a range of different flow cells, both low mid and high throughput to give that flexibility and that lower cost, lower experimental costs when you have fewer samples. 

 
46:34 
I guess that's always been the same, you know, you end up paying a little bit more per gigabase for that convenience. 

 
46:44 
But there is that growing flexibility, you know, to reflect I guess, what there is in the marketplace, what demands there are from the multiple users of different types of lab. 

 
47:02 
Yeah, totally. 

 
47:03 
Thank you, Mike. 

 
47:07 
Just to say, you know when I look at the comments, yes the Ion Torrent is fully integrated with the Genexus. 

 
47:17 
This has been applied successfully during the pandemic for many labs for sequencing the SARS-CoV-2 genome. 

 
47:26 
Another company that has fully integrated, you know, from extraction to sequencing is Complete Genomics, where you know they have a fully integrated system from extraction to sequencing production as well. 

 
47:48 
For the rest, you know, you can mix different systems. 

 
47:53 
Sure. 

 
47:54 
Super. 

 
47:55 
Thank you very much for your answers. 

 
48:02 
So I see there are some hands raised. 

 
48:04 
Tom Brown, I think you're the first one who raised the hand. 

 
48:16 
Sorry, can you hear me? 

 
48:19 
I wasn't aware I had raised my hand. 

 
48:21 
Sorry. 

 
48:22 
Interesting talk. 

 
48:23 
I did have a query. 

 
48:24 
I was trying to look up Jiannis’ paper and it seems I've hit a firewall. 

 
48:30 
Is that or is that because I'm not subscribed to the journal or is that because it's not available in the UK? 

 
48:37 
Oh, which paper? 

 
48:40 
The nanopore one, it looked quite interesting. 

 
48:42 
I'm working on something very similar at the moment. 

 
48:44 
So I was actually intrigued that you. 

 
48:46 
It looks very similar to what we're currently doing. 

 
48:49 
That was all. 

 
48:50 
I was quite the Lancet? The P6O6. 

 
48:55 
It might just be an organisational firewall here, but yeah. 

 
48:59 
I'll send it to you. 

 
49:00 
Yeah. 

 
49:00 
OK. 

 
49:01 
If you send me your e-mail. 

 
49:04 
I don't know. 

 
49:05 
I can't. 

 
49:06 
I don't know. 

 
49:06 
I wasn't aware of it. 

 
49:08 
I can pop it in the chat. 

 
49:09 
Yeah, if you don't mind, that'd be great. 

 
49:11 
Yeah. 

 
49:11 
All right. 

 
49:12 
Yeah. 

 
49:13 
Sorry. 

 
49:20 
And Alan Rico is your comment, you said you want to discuss this further? 

 
49:40 
Again, if anyone does have more questions, we do a few minutes left of this session. 

 
49:47 
So feel free to put those into the chat. 

 
49:49 
All right. 

 
49:57 
I'll go again. 

 
49:58 
If maybe also jumping in, you mentioned Pacific Biosystems and a benchtop. 

 
50:05 
Is that something you're aware of or is that something you're suspecting. That would be nice because it would give a bit of competition to nanopore. 

 
50:20 
Well, I cannot seem a lot, but it is real, OK. 

 
50:29 
It seems to be in the, you know, in the making. 

 
50:33 
Let's put it that way. 

 
50:35 
So I'll keep my eyes out then. 

 
50:38 
Thank you. 

 
50:41 
How do people feel about, you know, single cell technologies moving, you know, being translated into the clinic and the hospital? 

 
50:53 
Because if analysing, you know, genomic data, sequencing data is challenging, you can say that this is, you know, on another level, how to streamline single cell genomic data analysis and implement them into clinical practice. 

 
51:17 
Is there interest? 

 
51:21 
Give a more clinical angle to this. 

 
51:24 
I'd say that in order to be taken into in being useful, they need to be actionable, they need to change in some way the clinical protocol pathway process. 

 
51:43 
So it's all very good to know that one cancer is one type or another and that doesn't need to change the treatment protocol. 

 
51:57 
It is, well, it varies a lot between different types of cancers. 

 
52:04 
I mean with some cancers you'd probably, and cancer is like three decades ago for me, I kept an eye on that. 

 
52:14 
I was a tumour analyst. 

 
52:16 
Sometimes at some point, but in order. 

 
52:21 
Sometimes you don't really need to know the sequence in the next hour or so because you're going to do the surgery, remove as much as possible the tumour and then of course it's good to have some kind of answer soon. 

 
52:37 
But most cancer treatment pathways and protocols are pretty well defined. 

 
52:42 
So you may not need to have the next generation sequencing in the next two hours. 

 
52:50 
Moreover, with cancers cells being so heterogeneous, the question is, well, you're going to take a little bit from one end of the tumour, but it will definitely be different from the other end of the tumour. 

 
53:03 
So yeah, in order to bring it to clinic, it must change the treatment or at least the treatment protocol pathway in a significant way. 

 
53:30 
Yeah, I mean surely there are disadvantages over single cell in the clinic availability of cells from some organs, some tissue areas. Also costs, you know, ideally a clinical test would be super high throughput and £1-2 where a single cell, you know expense, right. 

 
53:57 
So yeah, I'm kind of wondering whether you know, single cells been revolutionary in research, whether the main basis in finding those biomarkers given time we'll learn, I don't know, imaging techniques and various other PR based techniques that will be super quick and you know scalable and cheap. 

 
54:30 
It will, it may. 

 
54:35 
What could happen is that, you know, one can capitalise from the research projects that do utilise single cell RNA installation, sequencing, etcetera. 

 
54:50 
And it is possible to translate that, if we consider the new special genomics methodologies where you know, you can apply a panel for example to a tissue section and get fantastic data that are not going to interpret in that sense, you can focus on the markers that are critical. 

 
55:17 
So it may, if it finds way as a you know, and let's say at this point it may. 

 
55:29 
The results of the single cell work could be translated through special Transcriptomic and Genomic technologies Hospital using panel focus panels that do not require sequencing. 

 
55:45 
You can get the results through methods that just, you know, visualise the tissue and the pattern of key markers, diagnosis more test that you know, you can interrogate mutations even with these technologies in a special context. 

 
56:10 
So that can be one way to introduce this type of methodologies in the clinical practice. 

 
56:24 
Brilliant. 

 
56:25 
Well, thank you both. 

 
56:26 
I guess we do have about two minutes left. 

 
56:27 
I just want to maybe do one sort of last sweep, you know, did anyone else have any last comments- Jiannis and Victor? 

 
56:35 
Otherwise I'll close the session. 

 
56:36 
So we'll just give it 5 seconds in case somebody was typing anything. 

 
56:41 
All right, again, thank you very much both for your presentations and your time. 

 
56:44 
It's always much appreciated. 

 
56:47 
So yeah, thank you everyone for joining and thank you everyone who contributed in terms of, you know, joining the discussion as well. 

 
56:51 
So I thank Patrick and Michael as well. 

 
56:53 
It's always much appreciated. 

 
56:54 
And again, thank you Jiannis and Victor. 

 
56:56 
Your time is always appreciated and without you these sessions wouldn’t be possible. 

 
57:00 
So thank you both. 

 
57:03 
Just general housekeeping notes or general PSA. 

 
57:05 
So next up in the Omics series for Oxford Global. 

 
57:07 
So we do actually have our Spatial Biology UK Congress, which is happening in London on the 18th and 19th of March. 

 
57:14 
I believe registration is still open for this. 

 
57:16 
So if interested, feel free to click on the website and check that out. 

 
57:20 
And I think our next Omics science exchange will be on the topic of utilising spatial biology and AI in late-stage clinical development. 

 
57:31 
And that will be on the 10th of April as well. 

 
57:33 
So you might get an invitation to join that as well. 

 
57:36 
Otherwise, again, thank you, Victor and Jiannis and thank you for joining and have a great rest of your day. 

 
57:40 
So thank you everyone, take care. 

 
57:42 
Thank you. 

 
57:43 
Thanks for hosting us, Declan. 

 
57:44 
Thanks. 

 
57:45 
Have a good rest of the day. 

 
57:46 
Thanks very much. 

 
57:47 
Take care. 

 
57:47 
Bye. 

 
57:48 
Thank you. 

 
57:49 
Bye.