Advances in Genome Biology and Technology (AGBT)

*This transcript was generated using an AI-based transcription tool and may contain errors.

I hope everybody has had a good AGBT thus far, and we really thank you for choosing to spend the next hour with us. I’ll start us off, and then I will be joined with my colleagues, Johanna Whitaker, who leads our sequencing research and development team; Professor Edwin Cuppen from the Hartwig Medical Foundation, an early collaborator on the AXELIOS 1 platform. He will be followed by Gustaf Calberg, who leads our sequencing systems lifecycle, and then last but not least, Mark Kokoris—he’s the head of SBX technology, the founder of the technology, and by no means a stranger to this stage. I also hope that everybody has had a chance to look at the platform in our suite. AXELIOS 1 is a two-system platform, and the separation of chemistry and sequencing measurement has really enabled it to be designed for flexibility, scalability, and efficiency all the way to analysis. The first of these systems is the synthesis instrument that in parallel can process up to four independent library pools. You can also process them together in parallel fashion up to four, or you can do one, two, three, or four depending upon your throughput needs. The reagents and consumables associated with this platform are modular that actually allows batching flexibility based on your specific needs. Optimized protocols will be available at launch to support various methods as well as various sample types, and then we will continue to build from there. The Xpandomer synthesis process happens on a fixed surface of the Xpandomer chip. The input is your multiplex library pool, and the output is an Xpandomer per pool of those libraries that can be then sequenced downstream on the sequencer, and only four hours are required to generate up to four Xpandomers on the platform. The sequencing system can then process these Xpandomer pools in a simultaneous fashion. So each of those Xpandomer tubes that I showed on the previous slide is one run, and that is tunable from a few minutes to up to four hours on the sequencer, and up to four of those Xpandomer tubes can be then set up on the sequencer to run in a simultaneous fashion to allow multiple or high-throughput sequencing if your lab needs that. The analysis module that is integrated with the sequencer allows for flexible and accelerated analysis, and we’ll talk about that as well. When Johanna comes up, she’ll talk a little bit more about the various sequencing run modes in terms of those individual runs as well as queued runs. The sequencing system uses a sensor module with 8 million microwells, and as we have shown previously, a large number of these microwells are available for the Xpandomers to pass through and a readout that results in high throughput. And the structure of Xpandomers include these specific translocation control elements that very precisely control the movement of Xpandomer through the pores. This results in a very clean signal and at a very high rate of more than 500 million bases per second. The combination of this high speed with the clean signal as well as the large number of microwells on the chip allows for both high throughput as well as the fast speed that we have come to establish with the SBX technology. So in summary, each instrument has its consumables and reagents. Again as a reminder, when we talk about the AXELIOS 1 platform, it consists of the chemistry and the measurement module independently. Each of them have their own consumables and reagents for intuitive UI for loading. However, regardless of the throughput or application needs, you will have the same set of consumables. So there are no different kits based on cycles or output that you need—you need essentially the same reagents and consumables. And I know many of you have been waiting for pricing. We will talk about pricing in this presentation. What I need you to remember is when we talk about pricing and we talk about sequencing consumables, we are talking about consumables across chemistry as well as measurement. There are no other fees to be paid; this is an all-inclusive price that we will talk about. And towards the end of the presentation, we will also give you a little bit of a heads-up on the commercial timeline. So stay tuned.

One distinctive feature of SBX technology is the sequencing modes. So we have talked about this before—the duplex mode and the simplex mode. In the duplex mode, the parent strands are physically linked in the library prep process by this hairpin index that you see on the left-hand picture. This allows us to do intramolecular consensus to improve accuracy for applications where you need that high accuracy, for example, whole genome sequencing. But in several other applications, for example, counting applications, you don’t need such high accuracy; what you need is a very high throughput. And that’s where we can use the simplex mode, which is also capable of longer reads up to 1500 base pairs under appropriate sample and library prep conditions. The simplex mode is also very versatile, so you can use it for single-cell, spatial, bulk RNA, as well as target enrichment and proteomics, and we’ll talk a little bit about that today. The input requirements of these workflows are pretty consistent with the NGS methods that you already use in your labs today. As with other methods, the input is typically a function of your sample quality and quantity as well as the needs of your specific application, and it is no different for AXELIOS 1. So what we are showing you here today on the slide is what we have demonstrated so far and kind of represents an optimized workflow. But as AXELIOS 1 gets into the community and the community adopts the platform, we are also looking forward to seeing how you stretch the limits and show what else may be possible with the technology. So as we are talking about library prep, I want to give a little bit of a view into the duplex library prep in which you are adding that hairpin adapter. So the DNA preparation is actually quite similar to what you would expect in other DNA sequencing methods, for example. So you fragment your DNA, A-tailing, and clean up. The differences here is that two-step ligation process where you add the hairpin adapter, which contains your sample index, and then you link the two strands, and then the Y-adapters at the end of those strands, which contain the sequences so that the library can be attached to that XP chip. Based on your input, you would also have an amplification after you have created that molecule. Instead of traditional PCR, we use linear amplification, and the time for linear amplification, very similar to how you determine PCR cycles, will be dependent again upon the input quantity that you started with. So something with a standard SBX duplex that we typically use in our lab, for which Johanna will show you data, we use about 50 nanograms of input and approximately six hours of amplification time. However, based on your input, this amplification time can be adjusted, and you can also do amplification-free workflows if you were starting with a much higher input and go straight into quantification and pooling. So from an AXELIOS perspective, you will be providing library prep to enable both duplex and simplex workflows. Now keep in mind that simplex is a very broad category, and in some cases, the ecosystem vendors will be providing some or all portions of this workflow, for example, for single-cell RNA, and we’ll talk a little bit about that in the context of a collaboration we’re driving with 10X Genomics. But the various components will include the core library prep kit—this is based on our proven Kappa technology—as well as these adapters, the amplification kit, as also a quantification kit that will be compatible with Qubit systems. You have seen the workflow; it’s pretty similar to how other DNA library prep methods work, and so again very similar to our Kappa portfolio, we will be automating the SBX workflow—the library prep workflow—on our walk-away automation platform, AVENIO Edge, as well as third-party ecosystem providers like Beckman, Hamilton, and Tecan. And then Volta Labs is also looking into integrating SBXG on the Callisto platform. Speaking of analysis, the data generation rates we talked about are quite fast; we are doing real-time base calling on the station that is included with your sequencer module. For counting applications, you would then be able to get those FASTQs pretty much as soon as your run is finished and start your analysis.

For SBX duplex methods, you would also perform—you can also perform demux and intramolecular consensus to then drive those consensus reads FASTQs into your standalone analysis. We also will provide an option to perform mapping and alignment on the station. The value of on-station is to get these files pretty much as soon as the run is finished, and as a reminder, the run can be a few minutes to up to four hours. All these files are available in open format, so these can be leveraged for ecosystem analysis tools. But we will be also providing XOOS analysis tools, starting with demux and consensus, but also a variety of variant types. All these tools will be available for the community freely to be used, and many of these will also be open-sourced and hosted at GitHub. Last year in September, we released the first of our datasets on a seven-plex genome in a bottle sample along with the small variant calling with XOOS, and these are all available on our website, so you can download it from there. We also have a collaboration with Google Life Sciences, and there is a deep variant SBX model that is also available along with a white paper to show how the germline small variant calling can be performed using SBX duplex. This was the start. Over the coming months, we will also be releasing additional datasets across different whole genome sequencing and RNA applications with different types of samples as well as application notes, and we'll continue to iterate from there. Using the various methods and the protocols on the synthesis modules as I mentioned, we have been working very closely with the community. There’s a lot of internal work that has gone on to show duplex workflow in the context of different sample types, and Mark presented quite a few of this data last year at various conferences, as well as driving collaborations with our early access partners such as Hartwig Medical Foundation, and Edwin will speak more about his experiences later today, as well as with Broad Institute both on the rapid whole genome sequencing as well as single-cell RNA based on 10X Genomics products and bulk RNA with the Wellcome Sanger Institute. And we do estimate to continue to work with the ecosystem. An example of that is proteomics workflows in AXELIOS 1. This is a workflow where the simplex method really comes into play with the large amount of reads that can be generated in a pretty short time. As proof of principle, we worked—collaborated with three of the vendors in this area: the proximity extension assay with Olink, the proximity network assay with Pixelgen, as well as the NULISA technology from Alamar Biosciences. And what we did here was we took pre-made libraries using the commercial kits from these vendors and we optimized them for the SBX work—for AXELIOS 1 using a prototype workflow. We then took that data, down-sampled to the same number of reads across the existing platform and kind of plotted them, and you can see very high concordance on each of those platforms. Now this is a proof of principle, and what we are really excited to see is how we then continue to develop along these assays as well as participating with other ecosystem vendors. With that, I am going to invite Johanna Whitaker to give us a little bit more information on SBX duplex.

*This transcript was generated using an AI-based transcription tool and may contain errors.

[Music: Red Hot Chili Peppers - Can't Stop]

Thanks, Meethu. It’s great to be back at AGBT, and thank you for spending time with us over your lunch period today. So over the past year, we’ve been working closely with our world-class partners to really demonstrate successfully the capabilities of our SBX technology. Today really marks an evolution in our story. So we are transitioning now from really demonstrating the technology to really showing you what the product looks like. So to that—and specifically what we want to look at is what will the performance look like in your hands when the platform is at scale in terms of both the chemistry and manufacturing. So to this end, we’re going to take a closer look at the comprehensive performance of our SBX duplex workflow for whole genome sequencing.

So to give you a realistic view of performance at scale, what we really wanted to do is to accumulate a set of data from 65 four-hour runs. And we sequenced over 1,000 high-quality human genomes, and we pooled these genomes at a level of 16-plex. Now we’re running these runs—the data was also collected on systems in our near-final product configuration, where we ran across eight different platforms, and this was eight synthesizers and eight sequencers. And the runs were carried out by 10 operators using 12 different sensor modules and including sensor module reuse. Now these are not formal verification runs, but we think this will give you a good view of what the platform will look like in terms of its robustness when you scale the platform. So before I dive into the data, I just want to reintroduce this concept of a sequencing run queue as you can see here. So our sequencer is designed to allow four Xpandomer tubes to be loaded at a single time on the sequencer, and the system will automatically transition between runs. So a single tube of Xpandomer is run in a single sequencing run as shown here on the left, where you have a one-hour setup of the pore followed by up to four hours of sequencing. Now a full run queue will consist of four sequencing runs with four Xpandomers and will run up to 20 hours. Now you’ll have the flexibility to run how you would like to, whether you want to run a single run or whether you want to run in this configuration of an automatic run queue, and we’re going to show you data on both of those today. So to characterize performance, we’re going to show you three levels of analysis: the first will be from run to run, the second will be within this automated run queue where we have the four runs, and then finally we’ll show you some run queue to run queue performance. So let’s start here with the individual run performance. So we really wanted to take a closer look first and describe again what the base categories you’ll see in a typical SBXD run. And if you look at the table here on the right, the vast majority of bases that we achieve—the bases that we generate—are concordant duplex bases. So just to remind you, by looking at the left, these are bases that are generated when we read both strands of the DNA molecule and the base calls are a perfect match. These are our high-quality greater than Q38 bases. And these are really important because the yield of these bases will be the metric that we use to set specifications for our system performance. The second class of bases are our simplex bases, and again if you look at the bottom left, these are bases where we read one strand of the DNA molecule but not the other strand in a partial duplex read. These bases are our Q20 bases and are still useful bases to do mapping alignment. And finally, we have a very small percentage of discordant duplex bases at 1.3%, and these bases are again where we read both strands of the DNA molecule but the bases do not match. These are assigned a very low quality score, and these are filtered out from our analysis tools. So what you can see here, 72% average of concordant duplex bases is highly consistent across all 65 of our runs. Next we wanted to show you what is then our total yield that we generate per run. So here on the left, we’re again looking at that critical total concordant duplex bases, and across 65 runs, we’re achieving 2.33 terabases of pure concordant bases per run. And we have a distribution around that yield, as with other sequencing platforms as well, and our range goes from 1.8 terabases up to 2.7 terabases. Now a key driver of that performance is shown here in the middle, which is our predicted sequencing pores, and you can see across the 65 runs that we see a very high—over 7 million pores generated per run. We also are showing you our raw base accuracy reads here on the right, and you can see that we’re exceeding 99%, and this is base—this is of course prior to any computational consensus that we do to align the forward and the reverse strands. Next I want to show you the performance that we achieve within this automated run queue. So I’m going to show you two different run queues. The first is where we keep the library constant. So as you can see here, we take the same library pool, pool A, and we actually load the synthesizer with that library pool and then from that we generate four Xpandomers and then we sequence those across the four consecutive runs on the sequencer instrument.

And here if you look at the metrics here on the right, you can see that we achieve greater than—we achieve 40X median de-duped duplex coverage for each sample, and we have a median concordant duplex accuracy of greater than Q38. Our mean insert sizes range between 230 and 250, and we have good coverage uniformity. So you can see that the performance across this run queue—and again this is an automated run queue of 20 hours—you can see that performance is really consistent from run one to run four. The next run queue is more of a real-world scenario where we’re looking now at four different library pools that we’ve loaded into the synthesizer. So we again create four sets of Xpandomers and we run these in an automated run queue on the sequencer. And again showing you the same metrics that even with different library pools, you can see that we’re getting, you know, good strong performance across run one through run four. So the ability to pre-load your set of four Xpandomers in a single automated run queue really provides sort of a convenient walk-away workflow that really maximizes your daily throughput and minimizes your hands-on time. So we also wanted to look at variant calling consistency as well across the run queue. And so we did this on—we wanted to give you an example here in the same run queue where we looked at variant calling using our XOOS small variant caller, and this was evaluated on the Genome in a Bottle high concordance benchmarking region version 4.2.1. Now we down-sampled the data to 30X median de-duped duplex coverage so that we could compare the F1 scores across the different runs. And so we’re looking here at the Genome in a Bottle samples HG001 through HG007, and the colors represent that particular variant—or F1 score that was generated in that run. And you can see that for both SNV and indels that the consistency across the runs within that queue is very—is very good. So next I want to move to showing you the performance from run queue to run queue as well. So these are again sets of 20-hour runs of four runs each. And when you look at that, we’re showing two different sets of data here. We’re showing again the important total concordant duplex bases or the yield. And again you can see these are 52 out of the 65 runs that we ran in a four-queue configuration. And again you can see that there is some variability between the yield, but that our—we have a very robust baseline where all runs generated more than 1.8 terabases of data. And on the right, you can see we have a very nice high consistent pore count across all of these run queues as well. Now I want to switch gears a little bit and talk about sensor module reuse. We know this is a very high topic; there’s been a lot of anticipation to understand more about how much we can reuse our sensor module, so we’re going to share some of the performance today. Before I go there, I want to just describe a little bit of the sequencing run steps shown here on the right, of which there are four. The first is really initialization of the run. This is followed by run setup, which includes basically generating the lipid bilayer with the pore. Second is flowing in the Xpandomers and doing sequencing, and finally, we do a rigorous system clean or wash of the sensor module prior to the subsequent run. And this happens between run queues and also between runs. So you might ask, how much can you do? How many runs can a sensor module support? So here we are showing a representative sensor module and the performance over 20 runs. We ran these 20 runs that we happened to run in a single week, and we’re showing you the performance based on our predicted sequencing pores, also the total concordant duplex space, and our mean concordant duplex accuracy. And you can see that the performance is maintained from the very first run all the way to the 20th run. Now although in this experiment we ran all 20 runs within a week, the sensor module can last longer than a week, and for example, many of our early access sites have used the same sensor module successfully for up to two to three months. And this is a target for our product configuration.

So you also might ask, you know, this is a representative sensor module, what is the success rate of reuse out to 20? And so we’ve been doing some longevity studies, which are shown here. This is a recent one in again our near-final product configuration, and we take a look here. We had 16 sensor modules and 15 out of 16 robustly made it to 20 runs. We had one sensor module that we had to retire at run 19, and it didn’t meet our performance threshold. But the great thing about our sequencer system, it’s going to be designed to basically do a QC check before the start of any run, and it will tell you if you need to replace the sensor module and it will do that prior to the start of the run. So again what this data is really showing us here is that you can reuse the sensor module again out to 20. Now while sensor modules are really built for endurance, they don’t last forever. There are factors that eventually will lead to some degradation over time. But again the product intent configuration will be able to support 20. Now with any reusable technology, there may be concerns about what—again there could be possible read carryover from previous runs. That has been a question that’s been posed. So we have been looking at read carryover. And when we take a look at this experiment here, we took a previously used sensor module and we took a single—single-plex unique indexed—or unique barcoded sample and we ran that in a run queue here over run 1, 2, 3, and 4. So you can see we had a different barcode ID for each of those. And then after those runs, you can look at the data and look back to see if you can detect any of the barcodes or sequences that were generated from the previous runs. And it’s as you can see here in the experiment that we saw zero carryover at the start of the run queue, and between run queues, we see very negligible carryover, so at point—less than 0.0015%. And this is much lower—a much lower rate than is reported by other sequencing benchmark systems. So with that, I would like to draw a close to my section of the presentation today and really highlight the core metrics that you can achieve with our sequencing platform operating at scale, again in its near-final product configuration. So I’ve shown you that we have very consistent performance from run to run across a run queue and across different run queues. Each sensor module—we see more than 7 million pores. We see good strong coverage uniformity across the genome for our samples. We’ve also generated in a four-hour run again that 2.33 terabases of concordant duplex bases, with that range of 1.8 to 2.7 terabases, with a concordance duplex accuracy of greater than Q38. And again the sensor module reuse—you can use a typical sensor module here for 20 times with consistent performance and we see a very low carryover rate of 0.0015%. And so with that, I would like to thank you for your time and also want to give thanks to the cross-functional teams at Roche for all the dedication over the many years to really get the platform to where it is today. And with that, I would like to introduce Edwin Cuppen to the stage.

[Music: The Black Keys - Beautiful Day]

I could have lasted a bit longer this music. Okay. Johanna, thank you very much for this excellent overview of the technology. It’s now the question how this works out in a customer’s hands. So I’ll try to give the story on the technology from our perspective. At Hartwig Medical Foundation, we are doing whole genome sequencing of tumor samples with the goal to use it for cancer diagnostics in actually routine clinical settings as well as building up a database that generates data so that we can improve care for future patients. So that database is available for researchers. So our key question is—these are my disclaimers—the key question is: can we actually use the Roche AXELIOS system for this same purpose? And from our perspective, there are five criteria, five factors that do play a role in assessing that and comparing technologies. This of course holds true for every technology, also an alternative to what we’re using currently in routine. First, that’s lab logistics complexity. If you do diagnostics, you want to have things as simple as possible. Turnaround time—patients do not want to wait, so it should be done as fast as possible.

Throughput should be scalable, it should be flexible, so that’s an important factor. Well, and of course, without the right accuracy, no technology qualifies, so that’s an important factor. And of course, at the end of the day, we have to be able to pay for the test, so cost plays an important role as well. So let me go into these five different factors and start off with what a procedure looks like in a routine clinical setting. So there is more than just sequencing—so four hours' run times is fine and great—but there is of course more to it, and from a test perspective what we always consider, you start basically with the samples coming in, which in our case is a biopsy and a blood sample, from which we have to isolate the DNA, have to do the QC. So all of that is happening on the first day, and at the end of that day, we basically start the library prep. We do that overnight because we automate everything, that’s basically what you want to do as well in a routine setting. And the next part is the Xpandomer synthesis, pooling again on the next day, and then starting the sequencer, that can be as Johanna just mentioned, you can start multiple runs in one cycle. How does this compare to what we’re currently doing? Well, there are a few steps which are different from our current Illumina-based procedure. The library prep is a bit more—has a few more steps. So you have more cleanups in there. On the other hand, we have automated all of this on a Biomek platform, so that’s okay, we can deal with that. The next day we have as an additional step the Xpandomer generation, takes four hours, 50 minutes hands-on time, so that is not too complex. After that step, we'll do pooling and start a—sorry, the pooling is done before the XP synthesis, was an error in there. We basically prepare the sequencer and start the sequencer. And there it is an interesting part because the sequencer run could or can only last for four hours, and the data is actually already being mapped while the sequencer is running, while the data is produced. So 20 minutes after the run is finished, we have a mapped BAM and CRAM file and we can kick off our bioinformatics. So really we can get—by looking at these procedures, we can shave off at least a day from our current fastest time from sample to reporting, and we can do this in less than three days on this setup in principle. Then let’s look at throughput as well. Basically, if we look at what can be pooled in a single run, it’s 16 genomes. We do tumor-normal whole genome sequencing, so we sequence the tumor three times and the normal one. So we can fit four patients in a single run. Takes four hours, we can start four runs, so we can do 16 patients in one day. Currently, we’re doing—and if you would compare that to an Illumina setup, that is basically—if you compare basically what you can do on the whole machine if you run it full time, so you can basically run this cycle four times a week, you end up with about 256 genomes, 64 patients in our case that you could do in a week. That’s approximately the same that we’re currently achieving on a NovaSeq X Plus setup using 25B flow cells. So the throughput is quite similar of this machine for the application that we’re using currently in routine. This is an example of: does this really work out? We randomly selected 10 patient samples, and although you see variation in primary characteristics like duplication rates and these type of things in the raw data, we do see that you get the same median de-duplicated genome coverage for all of these 10 patients for the tumor and a normal when you put six tumor-normal pairs on an Illumina NovaSeq S4 flow cell, which runs for two days, as what you would get from four tumor-normal pairs on a Roche AXELIOS duplex run. But you can run more in the same time. So all of that works well. So in conclusion, I would say if you look at the first three steps: lab logistics complexity slightly more complex but easily mitigated by automation; turnaround time is faster; throughput is similar, but we have a bit more flexible batch size because we can start with four patients in a batch up to 16 in one day without any consequences of not having runs fully filled up, while for an Illumina setup, we have to fill the whole flow cell to get to a lowest cost price level. Okay, now the real question because accuracy, I think, is key if you go to diagnostics. So the rest of my presentation will focus on those aspects.

So in the past about half year, we did almost 120 patient samples, tumor-normal pairs, that were actually sequenced and analyzed in a routine diagnostic setting for a comprehensive cancer center in the Netherlands. So we have the base data and use that as the reference point for assessing the performance of Roche’s AXELIOS system. We sequence 90X coverage for the tumor, 30X for the blood control, and all steps are performed at Hartwig Medical Foundation. So all the data generation, all the sequencing, all the library prep, everything has been done in-house by Hartwig personnel. So this is really the customer experience that I'm presenting here. We use duplex sequencing for high-accuracy variant calling for this application, and the tumor types, well we sort of randomly selected them. So they’re quite average on what we see—they’re all solid metastatic cancer patients, but this is sort of a random selection of tumor types. So they have a representative way of how it does perform in different types of complexities in the genomic landscape. This is basically the data over time. There are almost 500 genomes in this graph, and I think this is just confirming what Johanna just showed as well. We see very consistent performance over time, over runs, actually always meeting the specs that we need to see on the relevant parameters for primary data generation. So meaning that we get very highly consistent performance and also in an automated library preparation setup and by scaling sequencing. So the next step is a bit more complex because if you start using software, you can see a lot of differences and they can depend on differences and settings that you make in the software. And especially if you use different tools, it’s very hard to get to the same parameters. So this is basically why we further developed our software suite tool. This is called Hartwig Widgets. It’s a set of tools for comprehensive cancer characterization that goes beyond single nucleotide variant calling, structure variant calling, copy numbers etc. It also includes things like HLA typing, neo-epitope predictions, homologous recombination deficiency, tissue of origin prediction etc. So there’s a whole suite of tool that we use in our diagnostic setting and also for building up our database as a rich database for researchers. So we worked in the past year very hard on making this package platform agnostic. And that’s the work done by Thomas, Charles and Peter. I think they did a brilliant job building this first component. It's called Redux. It does de-duplication and type of things, but it also makes the whole pipeline input material and sequencer agnostic. And it does that by re-calibrating with—by doing an empirical re-calibration of the error values that you get off the machines. So we basically look at what is actually the real error rate for SNVs in a certain context, for indels in a certain context etc. And by doing that, not only the base that is called has the same meaning, but also the quality values that go into the pipeline get the same meaning. Because by origin and by how they’re spit out by the sequencer, that is very technology specific. So that's basically the principle behind this tool. And this is basically one example because it not only makes the whole pipeline platform agnostic, but it also allows you to compare the raw quality values of a platform in an empirical setup. This is just for the small nucleotide variants, where we’re looking at the comparison between single nucleotide variation accuracy. And this is for the duplex bases in on Roche versus the highest quality bin in Illumina. So we basically took the highest quality bins for both, but this trick is done basically for every quality bin. There are a few things notable: one is that error rates on every platform differ quite a lot based on the sequence context. So the type of change that you see is quite variable, what the error rates of the different changes are, but also the sequence context is very impactful, the trinucleotide context. So in certain contexts, error rates are much higher than in other contexts.

The left and the right graph are representing the same data. The right one is maybe a bit more easy to understand. That’s basically every change that you can have in the 96 trinucleotide context. And there you basically see that these Phred scores range between 40 and 60 for the Roche duplex technology and for Illumina between 35 and 45. And you see that on average the raw read quality or error rate in the reads is much lower on Roche than on the Illumina platform. Of course, you have to do all of this for all the bins. So for completeness, the numbers are at the bottom; I will not explain that. Does that matter? Partially, yes and no. But we see for regular—for standard somatic variant calling, Q values above 30 are typically good enough. It’s only specific applications where higher quality values start playing a role. So I want to give that disclaimer as well. So we should not exaggerate what we’re seeing here. It’s important to see that the quality rates are very high and good enough—and appear good enough for what we want to do with it. How does it then work out? This is small variant detection and indel detection done in 115 patients. We stripped out the MSI samples because otherwise they would dominate every data point in the first graph. And we see very high somatic variant allele frequency correlation for both SNVs and indels. If you look at the tumor mutational burden or indel burden, we get an almost perfect correlation. And also the amount of private variants is what you expect, and most of the platform-unique variants are stochastic effects because of at the cut-off levels that we’re seeing. So this is the noise that you also see if you sequence a sample twice on the same platform. Same for structure variants, there are different tools in the package that basically deal with this. You get break-ends, deletions, small, big, or biggest, duplications etc. And for all of these categories, we see an almost perfect correlation. And if you add up all the structural variants, we get to the same counts for all the different categories in this case. Then basically other types of things that we’re doing—so it’s important to come to the right conclusions about variants to know the tumor purity and average ploidy. We get—well, there is one example where we came to a different conclusion whether there was a whole genome duplication event. That does not impact any of the downstream work, but meaning this is also fully concordant. Next graph, copy numbers. If you look at chromosome arms, aneuploidies, or high-resolution 100kb bin copy number alterations of every event in all these patients, we again get an almost perfect correlation for both measures in terms of copy number. This is summarizing data for a single patient: point mutations, copy numbers, and structural variants in a Circos plot, and by eye you actually will not be able to spot a difference between the larger, high-resolution picture of a cancer genome between the two technologies. So we can look at derived parameters and measures which are clinically relevant biomarkers: microsatellite instability, tumor mutational burden classification, but also more complex homologous recombination deficiency which underlies a machine learning algorithm that uses input from different characteristics from the tumor. We get to an almost perfect correlation. There is one discrepant homologous recombination deficiency classification that is deficient in Illumina but not in Roche. There are no known mutations in known homologous recombination deficiency genes, so we actually cannot really tell if this is a false positive or a false negative. If you continue this, tissue of origin prediction, it basically uses information from basically everything that we know of the tumor molecularly and using a machine learning algorithm for that as well to come to a prediction. We get 100% concordance. We come to the same conclusion about what tumor type it would be based on the genomic information. HLA typing, also important for certain tumor types, also for predicting which neo-epitopes are presented. Our toolset has a specific tool for this, for which we can do four-digit high-resolution HLA typing. And basically for HLA for class one, which is used mostly, we basically get an almost fully concordant classification. There are three discordants: one we basically are sure that our software made the wrong call on the Illumina data; the other two we cannot conclude on that. So if you then add this up and summarize, we see that the reporting concordance for somatic variants is almost 99%. If you add in all the other parameters etc., which are clinically relevant, we go up to 1,200 events that we checked and we get to 99% concordance. And that from a clinical perspective and validation perspective is extremely high from our experience. So accuracy, I would conclude: this is also clinical accuracy, clinical reported accuracy, it is the same. So then costs, moment you’ve all been waiting for. It was just announced by Mitu we’re going to be talking about costs. So have all your cameras ready, I got the privilege to do this. Nah, not. Sorry about that.

I want to just end by mentioning two things. Based from the data, the clinical data and our experiences, we do see that Roche AXELIOS is suited for doing the same thing, at least according to the first four parameters. So we’ll have to see what the fifth parameter is going to make my decision or my conclusion different. And the final slide is this one: I also want to mention that we’re a non-profit organization that tries to work as open as possible. So all the software that we used for the analysis, none of this is Roche-based but is also completely open-source and available in different formats. But also the data that I presented, everything in this presentation is available. It’s privacy-sensitive data, so it is through an access-controlled mechanism that you can request it, there is the barcode and the link you can find there. So in principle, you should be able to reproduce the result that I presented today as of next month. And with that, I would like to acknowledge the people—the lab work under the leadership of Ewart de Bruijn sitting here in front of me; Joep de Ligt data analysis lead, and of course very importantly, the algorithm development group in Australia led by Peter Priestley. And of course the team at Roche that supported this journey. Thank you very much.

*This transcript was generated using an AI-based transcription tool and may contain errors. 

[Music: Frank Sinatra - The Best Is Yet to Come]

Mark Kokoris: So I promised Josh Lauer I wouldn’t dance when I came up here as that wouldn’t be good for business. Now that the price is out there, I won’t go backwards on that. So, thank you all for coming today; really appreciate this.

So, I'll stick to the future part of this. So as we kind of rolled this out last year, people didn't really know what we were doing. So we thought it was really important to give a good grounding of the technology—what is SBX?—so we went out with some publications with the preprint, did a lot of introductory discussions about this, and then started rolling out applications and, as Gustav mentioned, kind of pushing into multiomics at some of the workshops. And then, of course, what does that mean for the bioinformatics?

I see John and Chen—there they are, right there—they both talked about this along with Mitu and others throughout the year: how do we deal with all this data and how do we keep progressing the technology? And we've had the honor and privilege to work with some really world-class groups. Hartwig—Edwin, thank you for all the things you've done; the Broad team—I see them all out there; and Sanger as well, to really get their perspective on SBX. I mean, we want to hear it; we want to hear the truth about where we're going and what you can do with SBX as unique and really take advantage of the capabilities, the unique capabilities of the technology.

So throughout the year, I'll just give a few snippets of this. We've been able to really advance the capabilities as we rolled out earlier in the year with some one-hour sequencing runs, just kind of showing the possibilities there. Edwin covered a lot of this in his slides as well, just how the technology continues to evolve throughout the year. And then the extraordinary throughput that Gustav alluded to with the simplex and what the capabilities are, how you can really rethink your sequencing workflows and what's possible. And I have yet to see anybody who's done the initial work on some of these simplex not say "Wow, that's a lot of data." That's been universal, I think, which is a good thing; that's what we're going for: being able to push the envelope of what's possible. So as we expanded into multiomics, we did what we're calling SBXDM, which using the TAPS+ chemistry, which was just great right out of the box. We were able to demonstrate concurrent detection of variants and methylation status; that was a real great dataset. You should check it out if you can go back and look at the presentation there. And as well as doing some of the longer-read haplotype phasing, expression phasing as well, all at the same time. And then building on that—and Aziz has presented on this as well at a couple of different conferences—talking about the isoform identification as well, and really thinking about blending some of these applications together. Now you can do longer, you can do methylation, you can do panels and start thinking about those types of things as the technology evolved. And so when thinking about flexibility and efficiency of operation, I mean, we wanted to show people—and it wasn't really showing off, it was just showing what's possible with SBX Fast—to see that how you can compress some of these prep technologies and workflows into being able to do a genome in under four hours.

And so it was just getting people to think about the possibilities, about rescaling and adjusting their workflows in the context of this new technology. And so when thinking about that into the future, thinking about the throughput scales and how those can be applied in the varying read lengths, simplex and duplex across a range of applications, we’re just showing some of the ones that we’ve worked on already, but there’s quite a few more that we’re already looking at as well. And just thinking about what that means for the future. And as I said many times, SBX was designed for the flexibility and performance with headroom to scale into the future.

So our intent is to continue to scale the technology, and it really starts where it began: with the chemistry. So at least that’s one key part of this: there are a lot of levers to pull on this chemistry, in terms of the Xpandomer structure, how we make XNTPs, all the protein engineering to improve lengths, increase length and accuracy, as well as the synthesis conditions and how we do the synthesis, adding efficiencies and yields there, and same on the sequencing side. Similarly on measurement, you can imagine the array that we currently have—you can make a much higher-density array of that and at the same time think about faster translocation rates by pulsing faster. And then you can imagine the extraordinary throughputs that you can add to that. At the same time, sensor module designs can be redone to even improve and leverage that better. And then having the very good problem of having to deal with a higher-speed data path to deal with all of the data. So this is a good thing as we advance into the future, being able to really push that envelope further. And then the bioinformatics side, as John and Chen and their teams will keep working on enhancements to real-time analysis capabilities of the instruments, as well as the additional applications and work they're doing on the XOOS modules and tools, so look—continue to advance that work as well.

And then going back to some of the founding thinking of SBX for me, at least, was workflow. You know, I wanted something that was a single molecule that was very efficient, that you can get closer to the sample with. I don't like steps. So I think if you can reduce steps and efficiencies, that's where we want to keep pushing the technology into the future, thinking again about combining workflows. And then one of the things I get a lot of questions on is read lengths. So one of the biggest things on our radar is to push that yield at length in that 2kb range and just push those yields up into the many, many billions, and we do think that that is one of our primary focuses as we move into the future, being able to push that scale up. And then, yeah, of course, when you have this really complex technology that nobody really thought could work, and this great array technology coming together, of course it's of our interest to think about protein sequencing. So I'll kind of leave it at that in terms of contemplating where we're going on that, but it is definitely of high interest to us as well.

So I can say this—started the SBX, I kind of invented it in 2007. It was the same time my daughter Nikki was a baby. And now she's in college, which makes me now think how old I am, right? That much time has gone by. But they kind of grew up together, both, and now SBX is making it out into the world, going to college. And it's like the enthusiasm that everyone at Roche and the team has for what we're doing and the excitement that we have as it goes out into the world and so excited to see how you all use SBX and advance your very important research and again I admire all the work you people do. So it's just exciting to see the next evolution. So and as Frank Sinatra says, "The best is yet to come." So, thank you all.

Mitu Chaudhary: Thanks, Mark. And so the last question that everybody was asking was: what's the release date? Not quite there yet, but a little bit of a teaser: our commercial teams are getting ready to take pre-orders soon, so you can contact your local sales representative for updates, and the first commercial shipments are expected in the summer. And we'll be talking a little bit more about more specific dates in the near future. I really want to thank everyone for your time. I know it's been a long one hour, but I hope you liked the presentation and we'll see you soon. Thank you.

*This transcript was generated using an AI-based transcription tool and may contain errors. 

All right, thank you everyone for spending the afternoon with us. Really excited to be here now and actually answer that one question that most of you have asked us ever since we introduced the technology one year ago. And that obviously is, why did you call it AXELIOS?

No, we’re going to talk pricing. I’m sorry, I just had to get that joke in. We are going to talk pricing a little bit and get you a little bit more insight into what our technology can do. Before we get there though, I think it’s important to just come back to this, and Mitu touched on this already: the versatility of our system is such that we have lots of different things that this technology can do. And the simplex and duplex read modes are really important for what we believe we can bring with regards to flexibility for this particular system. And I just want to reiterate one thing: this is all done with the same consumables. So this is not something where you’re changing the kit or you have to think about what you want to do. This is all the same consumables that you have to use with that. And that becomes important when we talk about and think about pricing as we move forward.

So Johanna walked us through the representative dataset that we have generated. And we want to use this because we want to make sure that we can create a transparent view on what it will cost in your hands to use this system to gain deeper insights. We talked about the 1.8 to 2.7 terabase of concordant duplex output in four hours. And that’s what we’re really baselining this on, and as I get into the pricing considerations, that’s what we’re going to be looking at. We used 16 samples to run this study, and those 16 samples—if you think about it—are at the lower end of this range. So that’s what we’re going to be talking a little bit more about. Obviously, this is a very high-quality dataset with Q38 quality scores or higher, and we had about 250 base pair read lengths. We’ve also shown you lots of data to showcase how this system brings that flexibility and versatility where you can set it up to do a single run in four hours or, if you so choose and you have a higher throughput, you can set up a run queue and then process up to 64 genomes in a single day. So anywhere from 16 samples all the way through to 64 in a single day, set up depending on what your throughput needs are. I want to touch just on what we’ve shown you before as well about simplex reads because they are very exciting for us and important. We have shown how we can generate simplex reads—we get a lot more output on this—in this mode, right? And with 175 base pair read lengths, we generate over 40 billion reads in a four-hour run. We’ve also shown you the capability of this system to go to longer and extend those reads and get even more biological insights depending on what you’re looking for.

Mark has presented before different studies where we’ve gone—where we targeted the average of 500 base pairs of read length and where we’ve seen multiple fragments way over a thousand base pairs of read length. And in this case, we generate 10 billion reads in four hours—an unprecedented throughput in such a short time frame. Now these are of a lower quality score as we’ve talked about before, but there are plenty of applications where this comes into play with regards to counting applications and also if you want to build a consensus accuracy, we can utilize this run mode. So this really for us brings that versatility of what the system is. But then so let’s answer that question that you’ve all had: what is the actual cost of this going to be? So based on the study that Johanna presented and what we showed, we want to show you what it will cost you at 16 samples per run. Now this will enable you to get to a 30X coverage genome. And the price point for this will be **$150 per genome**. Now there’s a couple of things that I want to call out here specifically. First of all, this is based on real-world performance. We’re trying to create a very transparent price point where you will understand now what it will look like in your hands. Some of you might look at that range of output and say "I’m going to load more"—well, then you’re going to have a different price point, a better one. But we want to show what it really means from a perspective of when you are running this in your own lab. Secondly, this is a list price. This is what customers will be able to achieve when they run this platform. We’re not talking about discounted pricing here; we’re not talking about multi-instrument purchases to get to a certain price point; we’re really trying to give you what this will look like as you set this up and run it. Our assumption is that you run 16 samples and that you use the sensor module for the 20 times reuse that Johanna spoke about before. And as she also said, we’re going to enable you to do that over multiple months, so it’s not something that’s going to have to be rushed through. But these are really important factors for us to think about for this particular system. Now I know that this is the number that most of you have been thinking about—what is it going to cost per genome? But as I said, we have this capability of running simplex runs on the same consumables. So what does that mean from a pricing perspective if we go back to that 40 billion read output of 175 base pairs? Well, since it is the same consumables and you’re not having to buy anything different, you can pretty quickly do that calculation. Your price point for a million reads using simplex is going to be as low as **6 cents per million reads**. Again, this is at list price. This is an unprecedented price point in the market, which we believe will enable opportunities to generate much more insights to different applications and biological insights over time.

For just as a reference point, I think this is significant for the market overall. The best price technology in the field right now will enable you to do this for 24 cents per million reads. So this is a quarter of the price of what you can currently achieve with any technology in the market. We really believe that this is going to have a huge impact on what people will want to do with AXELIOS and what this technology will bring to you as you get it into your hands. So we’ve shown you now two price points: first of all, the $150 genome price, which provides significant savings for most customers of what you currently are getting from the technologies in hand. We’ve also shown you the counting and the 6 cents per million read, which is completely unique for what AXELIOS can do. But we also know that a lot of you are interested in thinking about—there's a big interest in very deep sequencing of panels, trying to get into an understanding of, you know, how do I use this, especially where we use UMIs to try to get into high—you know, being able to identify a high level of sensitivity and specificity for those kind of applications. At ASHG, Mark presented our target enrichment approach to this. And what’s really interesting with our technology is we can tap into this capability of the high—the extreme amount of reads that we’re getting for simplex sequencing, meaning those 40 billion reads. And while I’m not going to go through any more details on how we do target enrichment today—if you haven’t seen the presentation, it is available on our website from ASHG in October where Mark was going through this—but I wanted to talk a little bit about just this cfDNA panels and what this capability of the system will really bring you. Now obviously there will be a variability here depending on what’s your size for your panel, what are the applications that you wanted to do, but to give a reference point at least for pricing: if we imagine a 1.1 megabase panel where we’re trying to sequence the 2000X coverage, you can load 96 samples in a single four-hour run and that would bring you to a price point of **$25 per sample**. So AXELIOS technology for us really opens up a whole new capability for you in the market. For us, the $150 genome is just a starting point. I think the versatility and the flexibility of this system—the different run modes that we have available—really sets us up for a future where we can do things that we haven’t been able to do. So switching gears a little bit and talking about what are the things that we are looking forward to as we move forward. First of all, we’ve talked a little bit about multiomics in the past. We’ve shown some datasets that combine the methylation and DNA. We’re going to continue to work on this; we think it’s a unique capability of AXELIOS and the way that we generate data to give you those deeper insights to help the—help you all to actually generate data in the multiomics setting and combining datasets to get more out of our samples. On top of that, we are looking at applications where the capabilities and the knowledge of Roche can really sort of start utilizing this power and where we can bring an end-to-end workflow based on AXELIOS technology to the market, and we’ll be looking more and more to that as we move forward. But then finally, we’ve also talked about this being an open system, how we want to create something that that the industry can build on as well, really to ensure that AXELIOS is accessible for all those different applications that you might be interested in. We’ve shown a lot of work on the data analysis side. Obviously, we are utilizing libraries from NVIDIA and components—libraries from Parabricks that NVIDIA makes and components from NVIDIA to be able to keep—to create this real-time analysis workflow that we’ve talked about. We have worked with Google and DeepVariant to ensure that their technology can use the SBX data and they’ve shown datasets of that and that’s going to be available for you as we launch. We’ve also worked with 10X on the utilization of Cell Ranger and Space Ranger and being able to actually manage the SBX data flow. What we haven’t talked so much about yet is what happens on the upfront side of things. So today I’m pleased to announce that we are now working on a collaboration with 10X, really to start creating unique solutions for AXELIOS for SBX data and to actually start bringing single-cell workflows that are and kits that are designed for our system and the way that we manage that data. So more to come on this in the future. We’re also expecting that we’re going to continue to drive these collaborations with the industry and you’ll see more of these announcements for us coming out.

Speaking about the future, it gives me great pleasure to welcome Mark Kokoris to the stage to talk a little bit about where we are taking this and where we are going to go from here.

*This transcript was generated using an AI-based transcription tool and may contain errors. It reflects the spoken content of a live webinar and has been lightly edited for clarity.

Great. Thank you, Gustav. And he pronounced my name correctly. That's awesome. So, it's off to a good start. Great, so I've been thinking about things a lot lately, reflecting on the past and one of the things that came to mind was the very first time I did sequencing. So, it was quite a long time ago. Sanger sequencing, manual, one base per lane, autorad, your P32 and all that. And I'm bringing this up because two things. One was I screwed up, made a mistake in the experiment and didn't get the gel transfer step right. And two, because I didn't want my boss to know about that - and Seattle team don't listen to this, this is not something to listen to - I came in the middle of the night to do the actual sequencing of it. This is the first time I'd ever done it before. And I'm sitting there doing the sequencing and thinking, man this is so cool. That somebody figured out how to do this, and I'd say the wonderment and the awe that I had at that moment, I think it's something I never forgot, never want to forget about, and something that I think that was borne very early on in my mind, something that I wanted to be able to do is to develop a new sequencing technology and have that impact. 

And so I’m honored to be here presenting at this conference. It's been a long time coming, 18 years since starting Stratos. So, I will thank everyone for coming, and also to say that I very much respect all the technologies that I've seen here, and all the things that people are doing. It's amazing, and I very much love hearing about all the ideas and things that people come up with. 

So SBX technology. So, like any sequencing technology, you have to be thinking about accuracy, throughput, read length, cost efficiency. Those are givens. I don't care if it's 1999 or 2029, you have to be thinking about those things. But for us, flexible operation was one of the things that, when Bob and I started meeting at Starbucks for six-hour sessions, was right at the top of the list. We envisioned the technology that can sequence up and down the throughput spectrum, on the same system. Four minutes, 40 minutes, four hour runs. We wanted to be able to do that. In fact, I've talked to people, I'm thinking 42nd runs for some of the runs that I've heard from people today. 

So that was top of mind, and we'll talk about it a bit through the presentation. This will be a faster version of the presentation we gave at the webinar, but it is recorded so you can listen to it get a little more detail. John Mannion covered a lot of the data structure stuff there as well, so I recommend people go look at that. 

So we'll touch on accuracy. The results I'm showing here for duplex whole genome sequencing, for HG001 reference sample. You get an idea of the duplex sequencing accuracy that we're getting, and we'll cover that a little bit. Throughput, we did some seven plex whole genome sequencing reference sample, one hour run. We got over 30x coverage in that one hour run, and that's about five billion duplex reads in one hour. So you're getting an idea of the kind of throughput we're able to get here. We'll talk about the trade-offs between duplex and simplex sequencing, in terms of read length and accuracy. There's trade-offs there, to give a little bit of an understanding about that.

Time to result. We're going to show really cool result, Sean will cover that for sample to VCF whole genome sequencing in less than seven hours, which is really, really cool. And a fun way to represent the flexibility of this technology. So something that we've wanted to do from day one. And then cost efficiency. We look at it as all elements of the technology you have to be thinking about. Whether it's the chemistry, the measurement, including reuse of the sensor module, those kind of things. The quality of the data you're seeing. Everything, every element of efficiency, you have to be considering when you're thinking about this. So top of mind, we are very comfortable with where we're going and going to be able to go on the cost efficiency. And one thing, you know, be nice to Gustav! I know everyone's asking these questions about price and everything, and just to know that we've had him wired up the last week. I think Josh had the idea to wire him up in case if it gets in his mind to even talk about pricing, you know that he's going to get a little bit of a jolt. So, you don't want to see that. I saw it during the test runs, it's not a pretty thing. So anyway, be nice to Gustav. 

OK, let's get into it. So, OK, what is this? This is the coming together of Stratos genomics chemistry, and the genia array technology. I like to think of two technologies that needed each other to bring out the best performance. We showed the systems yesterday, introduced those. You can go see them in the suite, but basically a separate synthesis system and separate sequencing system. As you can imagine, those coming together at some point in time as a single unit is definitely something on our mind. 

OK, so what's our approach? So, our strategy for efficiently sequencing DNA was to not sequence DNA. It was that simple. Instead, convert DNA using an enzymatic chemistry process into an expanded surrogate molecule to address the signal to noise challenges of single molecule measurement. This was our approach it involved really on the edge molecular protein engineering techniques to do this. No one had ever done anything like this before. And so that was our approach. But when you get on the other side of it, you'll see the problems that you solve on the measurement side, what you can do when you have a chemistry, when you’re measuring a molecule like this. 

OK, the building block for doing this is something called an expandable nucleotide triphosphate or X-NTP. Our thinking at the beginning was attach a tether onto this nucleotide at two positions. And we tried many positions, but alpha phosphate and heterocycle made a lot of sense in terms of the biochemistry that we wanted to do. And then have those separated by a cleavable bond, and we use a bridging nitrogen amidate linkage there, to essentially, once you've incorporated this X-NTP in the copying step, replication step, that you can then expand the structure and now you have a code that represents the base identity. 

As time went on, we added translocation control because we knew we were really wanting to go in a nanopore direction and this allows for clean, reproducible, synchronized modulation of the movement of this Xpandomer through a nanopore. And we also learned that most polymerases don't like X-NTP, so we built in enhancers to actually improve the incorporation of the X-NTPs. And the last thing is nomenclature. SSRTs are symmetrically-synthesized reporter tethers. That’s what we refer to in the structure. 

The other big piece to this is a polymerase. For us it was pretty clear, we needed something that had an open architecture. You see where those arrows are, you need somewhere for these massive tethers to go. So we engineered a polymerase that had that open architecture. It's a lesion bypass Dpo4 polymerase, not conventional, not what you would normally think. But we needed to have something that had somewhere for those tethers to go. And I can spend all day talking about these last two slides. My team could, and they're here and ask them a lot of questions that they'll answer the ones that they can. But just overall, we had a iterative process of surveilling, refining, combining and iterating thousands and thousands of polymerases directed towards being able to better incorporate X-NTPs, and ultimately we ended up with something that has about 10% amino acid change. 99.3% raw read accuracy for simplex. Now obviously the duplex is much, much higher accuracy than that, but that's our kind of raw read total accuracy. That includes the polymerase copying step, and measurement, that's all rolled up accuracy. We turned it into a processive polymerase, which it wasn't initially and strand-displacing. So a lot of big engineering that we were able to do there. 

The workflow. I showed the synthesis instrument below, or in the suite we've shown that. It's basically just liquid handling, guiding reagents to this fixed solid phase chip that runs through a pretty intuitive process of, you have an oligo attached to a surface, you hybridize a template to it, add that cocktail of reagents to make the Xpandomer,  cleave to expand, and then using an orthogonal cleavage, photo cleave to release. That's the process. It's pretty intuitive, and it's what we currently do, and what we'll do for any sequencing right now.

So a little bit about that translocation control. This is a big deal because again, this comes down to measurement efficiency. Everything is about efficiency. And so I show there, with the green arrow, that little triangle holding the Xpandomer in place. This allows us to get good statistical sampling of the measurement, and you can see the level, the G level there. You then do a quick voltage pulse, advance the Xpandomer to the next step, and you see what that level looks like and so on, and so forth. So, you get an idea. This is a way of modulating throughput. So, if you pulse faster, you can increase throughput. And there's a lot of buttons that you can keep pushing on this technology to keep pushing throughput up. So that's one of them.

So that box there that I'm showing, was a drawing that Bob and I drew in 2012. This is what we hoped to be able to do. We couldn't do it then, but that was what we were drawing up. This is an image from 2019, so it took us a while, but this is what we were able to actually do. And you get an idea that we hit it on that signal resolution of measurement. That's a single molecule trace that we show there. You get an idea of what homopolymers look like, that we have this deterministic movement of an Xpandomer through a pore. So you can actually use that to call homopolymers, and we’ll show data later on how that looks. Quite good, actually. And then that's the resolution, the signal resolution we get between the bases when you look at the level histogram. So we hit what we wanted. 

OK, now you're on one pore. That was the data on a single pore on an Axopatch. Now you take it to eight million, OK. This is where the Genia technology comes in. You see the eight million chip on a PCB there, and it's essentially combining microwells, electrodes, detection circuits, and the A to D conversion. And then you get a picture on the right showing a bilayer, holding the pore in, and showing an Xpandomer translocating. 

So the sequencer instrument also liquid handling, to delivery agents, to the sensor module which houses the chip. And it basically runs through a very, this is a simplified version of the sequencing process, but you initialize a run, you set up the run by forming the bilayer and inserting the pore. We do this for every single run. And this is an iterative process, as I said. Then you go in and sequence for however long you want to sequence, whether it's four minutes or 40 minutes, clean the system and then you can reuse. OK, that's the process. Very clean. When we've looked at this many, many times, no crossover samples. And it's a very efficient process. And a lot of people have tried to do the Jedi mind trick on, on how many reuses. This is something that we'll talk about more over the coming months, but something that still is in process here. But we're very, we feel very comfortable saying significant reuse on that. 

Another picture to show is the pore count. So, at time zero, we typically get around seven to 7.5 million pores at time zero. But you get an idea over time, you know pores drop off over time, and that's kind of a typical drop off that we would see. But all the throughput that we're generating here, all went through this type of drop off profile. So another point to optimize with the technology, OK.

So, this is a one hour run single pore on the eight million array. The top picture you see open channel, you see the Xpandomer traces for many, many Xpandomers running through over the hour. And you zoom in and see what a 1000mer Xpandomer looks like. And the signal resolution you get there. That gives you an idea when you look at all the reads in that top window, how much throughput we're getting. And then you zoom in even further, and you see those translated XP base calls. 

OK, flexible throughput setup. We'll talk about this a lot, but just to say that if you want to look at one sample, four samples, because the synthesizer does four syntheses at a time, however deeply you want to multiplex or not, we want to be able to support these workflows flexibly. 

The paper. So we did a seminal SBX paper taking the technology up through 2020. So we really wanted a foundational paper to say, this is how you make these molecules and so that's available now. It has finally come up for people to look at, encourage you to read it. About 60-70 pages, you can scan the code here and have a look at it. Download took a while, I haven’t actually been able to do it yet so. But anyway, so it's a good read and kind of gives you an idea of the things we had to do to make the technology work OK. 

So, on the library prep, or at least the comparison between duplex and simplex, you get the idea. Duplex sequencing is basically measuring both parent strands on a single Xpandomer molecule. So, we can get intramolecular consensus reads on that. So that's the basic of that. And it will essentially show data on somatic oncology, as well as some of the fast applications, Edwin will cover the somatic, and Sean on the rapid. And so we'll go through a little bit of that. 

Workflow, pretty simple workflow. And it's essentially ligating to create that hairpin Y adapter structure. I say simple. It took a while to get a very clean workflow, but it is something that is very automatable and robust. And so we developed that protocol as well as a linear amplification approach. And that was key, it really helps us out a lot on the homopolymer accuracy, and so we were able to do a linear amplification approach and then just feed it in through the normal workflow to make Xpandomers in sequence. So other than that, you really don't know the difference.

OK, a little bit about the data. So this is that 7-plex genome in a bottle run. One hour run, we got over five billion reads into one hour. You get an idea of the breakdown of the number of reads per sample. You get an idea of the insert length coverage in that 34 to 38x coverage, and then the F1 score. So very respectable F1 scores for both the machine learning approach that John and Chen's team have developed internally, as well as comparing to the deep variant F1 scores, which they were very helpful and, and looked at the data and they actually got better numbers than us. So, it's nice to see that validation of our data in someone else's hands. Q scores. Q39, 99.7% genome coverage at greater than 10x. 

OK. The next workflow is SBX-Fast, which is basically removing the amplification step, and then tightening up the library prep, and just reducing some of the synthesis time to try and go for a really fast turnaround time on sequencing. And then just a little bit of data on SBX-Fast. We ran these a little bit differently. We ran them as trios. But basically five, nearly five billion reads, same type of quality of results. In fact, even better because we didn't actually even do the amplification step here. So great results, something that we really only just started doing. 

The vision, I think of the workflow, we think we can get in that five and a half hour range from sample prep all the way through VCF variant calling. We're actually not too far off of that and like I said, we just got started, but Sean will go through that.

A little bit about the read structure details. As you can imagine, we see this is a kinetic process. So you're going to see ranges of length, either full length or some partial. We use all of that information. We use all the duplex information, but we also use those single simplex tails. Those are really helpful for mapping purposes as well. So we use all the information that we have, very efficient, and you see very high yielding approach.

Accuracy over length was another thing that we looked at. John put this up here and described it in the webinar. Pretty much straight line. We don't see any impact, I think nothing to phase here. This is just single molecules running through a nanopore. On the homopolymer side, we see for all homopolymer lengths less than 15, we see F1 scores greater than 99%. That's both for our internal algorithms and development, as well as the stuff at Google Deep Variant, so very respectable homopolymer accuracy there.

A little bit about the accelerated data processing pipeline. Essentially this is where we're moving towards. As you can imagine, that's one of the strengths of the technology is to be able to literally read and analyze the data as it comes off. So looking at base calling demux, intermolecular consensus being done online essentially in real time, and then mapping and alignment in near real time. So that's essentially the direction we're going. And again, this is a big strength of the technology to be able to get that data quickly and flexibly and be able to use it when you want it.

A little bit about the simplex side. On the right, there is pretty much just single stranded sequencing. So we're not getting that intermolecular consensus, but there's a lot of good applications that if you want Q20 type of reads and a really, really high throughput like probe-based methods, or RNA isoforms, and Aziz is going to cover both of those in his session. This is a great approach to do that. And just to give a little idea of the yields and throughputs on the left, we show a 10x Flex run that we did. So a one hour run and we got nearly 14 billion reads in one hour. And those are near perfect reads because we pulled out from the total run, which was quite a bit more than 14 billion, just to give an idea of the throughput of links in that range. 

And then on the right side, you get an idea of a little bit bigger library size, around 800mer. And essentially if you look at all reads greater than 400, we end up with nearly over four billion reads in one hour. Over that, in that nearly 700 base read length, 99.35 accuracy and about 2.7 terabases per hour. So pretty substantial throughput. And again, like I've said many times, only getting started. Most of this data, we think there's a lot more in the tank. So I don't want to say disappointed. We're really happy to get this data, but there's a lot more here. So I see every experiment that we do. So I can say there's a lot, a lot more here to be done.

OK, So the wheel here in terms of the applications, this is what's so exciting, having talked to all of you here and some of you here today and this week is getting ideas about what people want to do with this approach. I love it. And I want to be able to see where we can go into the future and things that we can leverage this flexible technology. It may take some learning of workflows how to better leverage the approach, but that's, that's great. I mean, I think this is the whole idea here is kind of change the way we think about sequencing.

So with that, I can say I'm very excited again and honored to be here to go over the technology. It's been a long time, a long time to get here. Had the benefit of working with some amazing people over the years. And as I’ve said many times, Roche has been extremely supportive in this endeavor to bring this technology forward. So, I want to thank everyone and I think, Edwin, I think you're up. So, thanks everyone.

*This transcript was generated using an AI-based transcription tool and may contain errors. It reflects the spoken content of a live webinar and has been lightly edited for clarity.

Sean Hofherr: Hello. Pretty packed house! So, really excited to talk about our experience at Broad Clinical Labs with the SBX technology. So Broad clinical labs, we’re kind of nested within Broad Institute. We sit at this really interesting intersection between genomic explorers, resource builders, and clinical translators. So, basically folks across the whole entire spectrum. So, who’s looking for answers and new technologies, those folks that are trying to build up large databases so that consortia have access to really like, a plethora of data. And then, how do we take that information and actually apply it to clinical problems?

Here are a couple of examples of some of the folks that we've been working with over the years, and they run the gambit. And you know, a lot of these names are familiar and I'm going to mostly be focusing on the clinical translator side for the purpose of this talk.

At Broad Institute, we also happen to be kind of uniquely suited to test out a bunch of different platforms. So, we typically get the platforms before anybody else, so when they're in alpha release. Groups will come to us, ask us to try it out, give some real-world experience. We have the ability to provide input in real time before the product gets launched, which is really cool.

In addition, we also have like quite a large set of sequencers that are being run at scale. So this is like our current production lab. We have a large set of NovaSeqXPluses, we have a large set of Revios, we have some Ultima systems and then we also have some mid-plex and smaller sequencers as well.

So when Roche came to us, I guess this was like in the fall, and we had the first conversation about SBX technology, we were really excited. So, a couple of the things that jumped out and Mark covered this in the very first slide. The first aspect is there's a lot of data that can get generated from the sequencer. So that's like really, really attractive. In addition, supposedly high accuracy. At the point we hadn't really tested anything and the data that we were seeing was coming directly from Roche, so, we had to take it with a grain of salt. And then the speed of the system was also like really, really interesting. So being able to run and sequence genomes as quickly as we're able to do it on the sequencer. That actually unlocks quite a lot of potential. And one of the biggest places that we immediately thought that we could unlock potential is actually using this technology for rapid whole genome in the NICU.

So, today in the NICU, and Stephen Kingsmore gave a talk yesterday, he gave actually two separate talks. This figure on the right-hand side actually comes from one of Stephen’s papers and his group, and Rady Children's, has been pushing the envelope on rapid genomes for the last probably decade or so. 

What they found, and this has been replicated in many other studies, is that there's a huge diagnostic rate in the NICU. And so, when every single child in the NICU ends up with a genome, you actually find answers in like greater than 50% of the cases. And even in the cases that you're not finding an answer for, there's actually information that gets transmitted that helps in the management of that child. 

And so that's really important for a couple of reasons. So, the first reason is obviously you're improving the care of somebody that is like super sick and they're young, really young children. So that's really important. And the other benefit is when you start introducing genome into the NICU, you end up with, those kids actually spend less time in the NICU. And so that means now there's all of a sudden more beds like these valuable beds in the NICU for other patients that are severely ill to come into the NICU. And then lastly, the really interesting thing is that like the health economics of this makes a lot of sense. From the hospital setting, the healthcare industry setting, because having an early diagnosis actually leads to shorter stays. Shorter stays means less money that the hospital is taking out of pocket to cover the cost of that patient's care. And it's also less money that insurers are paying. So overall, this is a really attractive target. It's been a really attractive target. It's gained a lot of interest recently and it's starting to pick up and become routine care.

So, when Roche first came to us, they had this assay and it worked and we were kind of looking at it. And it was able to sequence a set of genomes in a certain amount of time, which was very nice. We wanted to know, how could we actually like dial down each one of the steps? So, as Mark said, there's a lot of levers. So you could decrease library prep time, you could decrease Xpandomer synthesis time. The sequencing itself is a big dial. So, you could actually have shorter sequencing runs and maybe load less samples on, and so you can get it as fast as possible. And then actually, how do we get variant calling to be as expedited as possible? Each one of these steps together there's a lot of opportunity to really shave off a lot of time and get this as fast as possible.

So, we got the instrument installed in January. So our colleagues actually had a little bit of a head start on us. And this is mainly because we were moving our facility from one building into another with all those sequencers. You could imagine how much time that was. So, in January, after the holidays, we came back, we had to install the system and actually within like one day, we were generating data. And so this right here, I'm just showing some of the initial data that we ran on the sequencer. So everything I'm going to be showing was generated in our laboratory, by our people. All the analysis was performed in our hands, and all the data came from us. 

And so, you know, the coverage is definitely higher than one needs for a genome, but this was kind of like the first shot. And so, we were just getting our hands, to get used to the system, and actually start thinking of those levers and how do we dial them up and down. But overall, the F1 scores for SNVs and INDELs were extremely impressive. And so right away we started thinking that, you know, maybe this promise that Roche came to us with, I don't know, six months earlier actually has some like some reality to it, which is really exciting.

When we start looking through all of the variants and we see how performance is across different variant types, interestingly, you know, this is a nanopore system and as you can see here, homopolymer performance is fantastic across precision and recall. And so this is like really, really impressive to come out of this system with such good homopolymer accuracy.

In addition, HighGC, LowGC, really good performance, really strong overall. Similar patterns for INDELs. I'm not going to go and kind of point out the differences, but in general the performance and profile is really, really solid for these variants.

And so, benchmarking with HG001 through HG004 is it's great, it's good, everybody's got to do it. But we wanted to know, are we able to find causative variants in individuals that have been diagnosed with genetic conditions? And so, we started testing some different variant types, so, cases with different variant types. The first set of samples I'm going to be showing are coming from just Coriell cell lines. Most of these are, are actual patients that were impacted with conditions that would be typically diagnosed in the neonatal setting. 

This first case is a case patient has propionic acidemia, it’s an inborn error of metabolism. These patients, actually in the first couple of days of life, if undiagnosed and untreated, even with treatment, they typically have like hyperammonemia episodes and they crash metabolically really early. And so, these kids do come to the NICU. And so, you can imagine just being able to have an answer for these patients could actually influence the care, because dietary management is a part of the treatment for these patients. And so, this is a patient, this patient has a single base deletion in the PCCA gene and we were actually able to detect it without any issue. This patient is homozygous. This is part of a trio from Coriell, so we do see it as heterozygous in both parents, and I'll show that data in a little bit. 

We wanted to look at performance across copy number variants. So we don't really just care about SNVs and INDELs. We also wanted to know how are we able to pick up copy number variants. This is a patient also Coriell cell line Duchenne Muscular Dystrophy. This particular patient has a 28kb duplication that actually like duplicates a couple of the exons in the DMD gene. It leads to Duchenne Muscular Dystrophy, genes also associated with Becker Muscular Dystrophy.

We were able to pick this variant up without any issues using the platform. Going away from the standard traditional types of variants, we're also interested in STR. So repeat expansions are a big interest for us. And so, the first patient that we tested, so that once again Coriell cell line, but we looked at a patient that had a full mutation in FMR1, so clinically affected with Fragile X. This patient we were able to run and detect the variant in this case.

These were not the only cases that we actually ran. We looked through a whole set of Coriell cell lines with different variant types. So you can see the first cluster, all small variants, they were all detected. The next set are all copy number variants also detected. And then the last set are repeat expansions with full mutations. And so, in each of these cases, the causative variant was identified. 

And so just thinking about the continuum, Coriell, great. But how does this actually work with patients? And so Broad Clinical Labs, we’re CAP accredited, CLIA laboratory also New York State accredited and licensed, so we have patient samples in house that have been diagnosed in our laboratory. And when we looked at those samples and we reran them, these are patients that have been consented to have additional testing performed with residual samples. You're able to see here that this set of samples, we were able to detect the causative variant that we actually reported in our laboratory for each one of these patients, which is really cool.

It's OK to be able to do this and see the variant in a VCF file or an IGV, but that's not really how we analyze cases in the real world in the clinical setting. So typically we take a phenotype driven approach and it goes into a tertiary analysis platform. This is an image from Fabric Genomics, Fabric enterprise platform. And what you see is that this is that PCCA, propionic acidemia trio and the causative variant was able to be actually prioritized using the AI. And this was without any training. So, we ran it as if it was a short read genome, threw the data in without actually even telling Fabric, and it worked perfectly without any issue and was able to find the causative variant. And this is what it would have looked like if we took this through a report. Just note this was not used for clinical purposes. We were just testing out to see how we could do it.

So, we got a little bit ambitious as we were going through this process and we thought, we are probably going to be really fast. And when you start looking, there's actually been a couple of Guinness records. So, Kingsmore had the Guinness record in 2018 for 19.5 hours, and this was using Illumina short read sequencing. Later he was able to cut it down, his group was, to 13.5 hours. Really, really impressive. In 2022, Euan Ashley's group using ONT technology was able to actually get a genome sequenced in five hours and two minutes, which is crazy fast. And actually going from patient to diagnosis, in a little bit over 7 hours. 

And so we actually like saw this and thought it was a great opportunity. And just without any training, or really much effort, we were able to get it down into six hours and 25 minutes. And so, we've only had this instrument for about six weeks. And so really the potential is huge and we're really excited to see how low we can go.

This was the work of a lot of people. Here are the main players that have been involved in this project for the last couple of months, and that's it. Thank you.

*This transcript was generated using an AI-based transcription tool and may contain errors. It reflects the spoken content of a live webinar and has been lightly edited for clarity.

Edwin Cuppen: Thank you Mark for warming up the room. Yeah, so the biggest question of course is for Gustav that everyone has in his mind. But the other question of course is also, how does all of this work in the wild? So, what does it work in a routine lab, and  what can you already deliver with that?

And I'm very excited to be presenting for the first time the work that we did in the past, well, less than six months. First a brief introduction on Hartwig Medical Foundation, that is a nonprofit organization and we’re actually doing routine whole genome sequencing based cancer diagnostics for various hospitals in the Netherlands. Besides that, we've actually developed open source analysis software to basically analyze this genomics data to drive data driven decision support tools in the hospital for oncology.

So these tools are used for two reasons. One is to generate a patient report with therapy guidance that goes back to the hospital. And the other part, with the consent of the patient, is to collect this information and bridge it with clinical metadata. And that is to basically build a Learning Care Cycle where with the data of the patients of today, we improve care for tomorrow's patients. So that's the objective of our foundation.

We do this whole process under ISO accreditation, and our average turnaround time is about six days. What is hampering us to get towards routine implementation, of course it’s new technology and it's a lot of data, but typically we're hampered by two main things. And one is basically this is too costly and it takes still too long. So, this is why we're always excited to explore new technologies to see if we can address these challenges. 

So we are very happy that we could team up with Roche on their new sequencing technology. This is a bit of timeline of the beta test that we did. You basically see the prototype equipment there, not as nice and shiny, but nice orange, which is a Dutch national color for the Xpandomer generation. So basically in October, November. So, this is installed in September 2024. So we started off with some test runs and optimization of our software, because I do want to stress that we try to be as independent as possible from Roche. Especially for the data interpretation, because we feel every new technology needs to be assessed under high scrutiny to see whether it's on par with what we're able to do already in our controlled clinical accredited environment.

Actually we did the first prediction run in December, and actually what I will be talking about is work and data that we actually generated at the end of January, which is done on 30 patients and some COLO duplicates. Ewart and Joris, I want to acknowledge them specifically because they really got the equipment very quickly working and under control in the lab. And also, Thomas and Peter, very instrumental as well to basically adapt our software to be able to work with data coming from the new sequencing equipment.

I don't have to say anything about this whole process. Mark already explained that. I will talk about 30 tumor normal pairs and two COLO control duplicate. Be aware, we do tumor normal pairs and we sequence actually per sample, per patient we sequence four genome equivalent. So it is a whole experiment and compares in total more than 120, 30x genome equivalents. So that was spread over eight, four hour runs that was done on a single chip and the sequencing was done over a couple of days to generate this data.

This was still done in a manual way, but Ewart and Joris also succeeded in automating this whole protocol and going from sample to data analysis or sorry, to data production, on four samples that could be done in 24 hours in an automated setup on Biomek. So meaning this is a versatile protocol that can be easily automated.

OK, what do we then get? Let's walk you through the different steps from raw data to the complete analysis. This is basically what you typically get, so what you see in these graphs is the tumor normal data and this aggregate information of the 32 samples. So on average, we've got 150 gigabases of data for normal, and 410 gigabases for tumor.

The spread of coverage is very good, meaning that there's not much variation between samples. This coverage, this is duplex data, but it's derived from two different setups. Mark just mentioned that full duplexes and partial duplexes in our setup. On average the duplex rate, the full duplex rate is about 70%, and 50 for these full duplexes is about 190 base pairs.

The partial duplexes have a slightly longer insert length, but you see that the termination of the duplex part is sort of random over the stem and that basically results in a duplex, average duplex length N50 and 110 base pairs. If you then convert it to mapped coverage of these reads. So that was the raw data, this is now the mapped data. We basically see that we get about 105x coverage for the tumor, and 47x coverage for the normal sample.

For the rest of the presentation, I will keep comparing this with our production data for these samples. So these are patients that were also analyzed on our inner production setup, that is currently running a Novaseq6000 and these patients had identical tumor coverage and slightly lower normal coverage.

OK, let's just look at coverage. GC biases. I think, don't look at the height of the curves, just coverage differences. But I think you should focus on the left and the right of these curves where you basically see that there is differences in GC biases where you basically see SBX coverage dropping off earlier than Illumina at low GC, but at high GC the performance of SBX is much better than Illumina.

So this results of course in dropouts in which you can nicely see in IGV views, you see dropouts in SBX in AT-rich regions, but the same thing happening for Lumina in GC-rich regions. If you look at the overall numbers the coverage, genome wide, is not much different. There is of course a second effect and that's the reads are longer. Compared to 2x 150 you have 300 base pairs to map something in a genome, and also you have a mate which is slightly further away from the other sequence, the other mate, so you get lower mappability due to the shorter reads. 

You see that in this in the white parts, the white reads have a map quality of zero, so they can also match somewhere else. You actually see that this region in this screenshot here already has more challenges in Illumina data as well, but you see that due to the shorter reads this region start dropping out completely with SBX data. However, if you look on where this is happening, and it's mainly outside the genome in a bottle, high confidence regions. So, if you correct for that and you just look at high confidence regions, you basically see almost identical performance for both platforms in terms of coverage.

OK. That's coverage, that's the basics, that's where you start. The next step is basically variant calling. We did this with our package of tools. So this is a package that we use in routine diagnostics. And it's not just variant calling it, there's also HLA typing, viral detection, all these types of things that can be relevant for the clinic, and for a cancer patient. You see the features here, you can use all the different types of data as input from FFPE, from panel etcetera. But one of the things how we designed this software is also to make it platform agnostic. So it should run with every platform. And the key tool in this is REDUX, and that's a pipeline component that deals with platform and sample specific characteristics and that feeds into the small and structural variant calling tools. And we basically took, well, took us a few months basically to optimize this tool to also work with the SBX data which has its own characteristics. And the rest of the tools of this pipeline were not modified, meaning they should all work once you have the REDUX tool properly working for this SBX data. 

So what this REDUX specifically does, it actually determines empirically the error rates to enable this high sensitive somatic SNV and INDEL detection. So that’s not the machine learning that you do once, it does that for every sample that comes in. So an FFPE sample has a different error profile and that's taken into account for the variant calling. And this is basically what it looks like for SNVs. It's trinucleotide context error rate that we determine based on a few percent of the genome that we survey for errors compared to normal reads. And this actually allows us also to calculate the empirical error rate from the different platforms.

On the SBX platform there is no different quality bins like there is for Illumina, but you have simplex and duplex reads. And if you apply this tool, so this is completely independent of any of the tools that Roche uses, we actually find the Phred score for simplex, for SNVs of 30, and for the duplex reads it goes up to 50. And if you basically compare the two platforms and consider all the data, we actually do see that the empirical platform SNV error rates are lower than on the Illumina platform.

Doesn't matter, this 1.7-fold or whatever way you calculate it. I'm not going to argue about that or make pleas about that. It's eventually. I think everything above Q30 is basically suited for the way we do variant calling in this for somatic variant identification.

OK, now basically what did we find? This is basically a COLO sample and basically here you see the challenge that you have with sematic variant calling. It's not a bin of zero, 50%, 100% variant allele frequency. No, it goes all over the range from zero to 100% and basically at the bottom of the graph on the left corner, that's basically where noise starts meeting real signals. Actually it's mixing, it's not even meeting each other. It's basically a mixed part of signal. So, it's difficult to interpret these graphs. Basically what you see in this case with this the raw data, you see high degree of shared variants.

For where do the private variants come from? Well about 20% is along these axes, no coverage in one or the other platform. That explains already a big chunk of the private. And the other part is basically if you look at the private variants and the variant allele frequency in the other platform, you basically see that they're all very low variant allele frequency. So we're actually looking at stochastic effects around just meeting the threshold for variant calling. So a simple thing to get this out of the equation is just to make a cut off of 10% variant allele frequency and then do the comparison. And then you basically see that you get sort of two identical numbers with the sort of symmetrical private profile, again around new cutoff that we defined. We believe these private variants are real variants when you look at their mutation spectrum.

This is about patients. You see the same pattern and you see very high concordance between a number of variants that are called in these patients with very low rates of private variants across these 30 patients aggregated. INDELS, same story, we basically see well, this is the COLO again. Please note that there is only very few INDELS in COLO. This is the melanoma sample, but the results are very clear that you don't get high rates of noise because of different specific error profiles.

Patients, same story for the INDELs. A lot of TF variants, and very few privates. The only privates that we do see are happening mainly at T homopolymers, which you basically see here. T homopolymers have much more abundant than C homopolymers in the genome. You basically see, above length of six or seven, you start seeing more Illumina private variants. We don't know if they're false negatives in SBX or false positives in Illumina. I just want to make a note that we don't use PCR Free Library prep in this setup, so these could also very well be PCR artifacts. 

Structure variance, copy number alterations. This is still work in progress. This is the first screenshot on COLO829. You basically see the inner circles that the copy number profiles are highly similar from both platforms. Structure variance is also looking quite promising, but we need a bit more tool optimization to get this completely done for all the patients as well.

What is the impact of private variants? Actually, if you nail this down, we see private variants across these 30 patients. But if you start looking at how many are in exoms and how many are actually actionable genes, clinically relevant genes, there are only 5 variants left, and they're spread over the two technologies. And the reason for missing them in one or the other technology is actually quite a variable, as you can see on this slide. And we basically come to the main conclusion that no clinical impact, there was actually no clinical impact due to platform differences.

I'll skip over the summary because this would be repeating what I just said and just come to the end conclusion that we feel that that's the SBX provides a very promising alternative platform for rapid whole genome sequencing based cancer diagnostics. Thank you for your attention.

*This transcript was generated using an AI-based transcription tool and may contain errors. It reflects the spoken content of a live webinar and has been lightly edited for clarity.

Aziz Al’Khafaji: Alright, good afternoon, everyone. It's a pleasure to be presenting at Roche’s workshop and share some of the efforts that we've embarked on with Mark and team over the past few months.

My name is Aziz Al’Khafaji. I'm the Director of Molecular R&D at Broad Clinical Labs and I lead the methods development lab, where we focus on creating and deploying cutting edge genomics tools.

When Roche reached out with the specs of the new SBX platform, we were excited to deploy it in ambitious ways that could mark a paradigm shift for key activities that we have going on at the Broad. We identified two main areas in the RNA space. The first is high-scale multiomic drug response profiling in multiplexed prism cell lines, and the second being ultra-high throughput single-cell RNA isoform sequencing with SBX. My main focus for this presentation will be on the single cell project, which Houlin and Guoping have led, with support from Brian. 

So just some very brief background. KRAS is a commonly mutated cancer driver and was previously considered an undruggable target, until researchers developed small molecule inhibitors which could inhibit that aberrant signaling. Despite very promising clinical outcomes, therapeutic evasion, drug resistance is still a big challenge. And if we can actually understand these routes, these mechanisms of evasion, we can actually drug those, have combination therapies, and realize the immense kind of potential within these drugs. So, rescue that potential.

To tackle this problem, we partnered with the PRISM team at Broad which has developed multiplex cancer cell lines consisting of 500 unique cocultured lines which can be distinguished through sequencing of their Express Prism barcode. This is an incredible resource which allows us to efficiently profile drug effects across diverse cancers in a single experiment.

So, we treated the PRISM pools with our RAS inhibitors BI-2865 and RMC-6236, along with our negative control DMSO, and a positive control that elicits a strong transcriptional response. So panobinostat, this is an HDAC inhibitor. We performed single cell at the three and six hour time points, and then as well a survival assay after five days to understand, you know, the survival outcomes.

With this experimental setup, we were able to actually link survival outcomes to early and intermediate kind of drug response. And the idea being, can we find these transient key intermediate states that allow for that ultimate survival, and those are potential targets that we can validate. And if proven true that we can go after them with secondary compounds.

And so, our primary readouts are the PRISM barcode. Again, this helps us assign and identify the specific cell line that the transcriptional profile is part of. We also, you know, this is a multiomic, single cell profiling. And so, we actually captured protein expression through the protein text discovery panel, which gives us reports on greater than 320 protein expression targets. This is in the context of the 10x Flex single cell platform which is a probe based, whole transcriptome read out on the Gem-X. And of course we're sequencing this on the SBX platform which gives us, for this experiment, 80 billion reads.

The data was phenomenal. I believe it was Mark who said that this picture looks like a picture from the Hubble telescope, and indeed this is a universe of data to explore. So I'm just going to go through some high-level metrics that we're all used to looking at with single cell. So here we had a median of about 1000 cells per cell line. For the per cell metrics we got about 28,000 UMIs per cell, 7,000 genes per cell. So the data was again very high quality.

There's no obvious batch effect detected. There's actually no green pixels on this image. This is actually the kind of co-clustering of our two different channels from single cells. So orange and blue kind of make green in a visual cortex.

Just as a high level sanity check, cell lines from the same tissues are more transcriptionally similar. So you can see on the bottom left the skin cancer cell lines, the top colon cancer and kidney cancer, etcetera. We also pseudo-bulked the different cancer cell lines from the single cell data, and compared that to orthogonally derived bulk RNASeq and they're highly correlated. So, we just have very strong confidence in the quality of this single cell RNAseq data.

Equally so, the protein expression data was very impressive. We were able to resolve many cell lines from protein expression data alone, which is pretty radical. So for the metrics there we see almost 13,000 protein UMIs per cell, and a median of 300 proteins detected per cell.

So, this slide underscores the utility of capturing both RNA and protein. So, some RNA protein expression is highly correlated, as you can see on the top right graph, and then on the bottom right graph some of it is poorly so.

So, each dot in that graph is a cell line. The X axis is RNA expression, the Y axis is protein expression from the panel. So you can see for a housekeeping like gene, it's highly concordant, right? But for things like transcription factors that are highly post translationally regulated, you know, i.e., protein degradation. If those switches are turned on and off, you're going to see a lot of difference there. 

So RNA is not always a good proxy for what's going on at the protein level, especially transcription factors. Obviously very important. So, this just highlights the power of capturing both of these modalities. The graph on the left just shows the range of correlation expression across cell lines. And so there's a lot of kind of, you know, that variance.

OK, let's dive into to drug response. As expected, we observed a range of transcriptional responses to our inhibitors. So annotating cells by treatment, we can use colocalization to get just a very high level, first glimpse at the various responses. Here we can see that the treatments are not colocalized with our negative control DMSO which is in purple, and this shows us that this is an example of a large transcriptional response to the inhibitors. Here DMSO and RMC treatments are colocalized so pointing to less of an effect on this cell line with the RMC drug. Vice versa for the BI drug, and here's a case where it looks like this line was not responding to either inhibitor.

OK, shifting focus now to the five-day survival data, where dark blue is sensitive and yellow denotes more resistant cell lines. I'll highlight four main classes that these cell lines fall into. Starting on the very left, this is resistant, cell lines that are resistant or sorry, sensitive, to both compounds. Just to the right, sensitive to RMC, furthest to the right, sensitive to BI only, and then in the middle resistance to both right.

In the interest of time, I'll just show that there are unique enrichments of mutation in these different classes, many of which are concordant with the literature. So, in the sensitive to both category, we can see that KRAS is actually enriched for, which is entirely in line with the mechanism of the drug. These drugs target mutated KRAS, and then we see different enrichments for various mutations for the uniquely sensitive cell lines, and then for the resistance of both, you can see that BRAF is one of the main enriched mutations, which makes total sense. This sits downstream of KRAS and does not need kind of this aberrant signaling from a mutated RAS to have its oncogenic effects.

So, this data was produced a little more than a month ago. So of course, this is just the beginning. From this first look, it's clear that there's a lot to discover from this data set. And we're excited to bring to bear, you know, these classical analysis, advanced AI approaches, and we will surface some of the complex biological signals.

And you know, we feel like this is a high yield approach. To systematically identify mechanisms of resistance, to be able to validate those, target them. And you know, I think really just mark a paradigm shift for drug development.

OK, great. Now I'll shift gears, and we'll briefly touch on our efforts in the RNA isoform sequencing space, with SBX. This work was led by Christophe, Brian and Houlin. So, we ran kind of like a traditional assay PBMCs and a 10x 5’ gene expression kit, and we sequenced the pre-fragmented folding cDNA libraries. We were able to resolve the canonical cell types and their biomarker genes, everything was kind of nicely clustered. But we wanted to actually dive in and, and see, you know, what these read lengths look like and their ability to resolve isoforms. So, on the far right is the read length distribution, the lion's share of which is between 500 and 800 base pairs. And you can see, you know, a fairly long tail heading to the right there.

To understand the impact of read length on our ability to resolve isoforms, we informatically binned those reads from 300 to 1200 in hundred base pair increments, and then we ran this kind of isoform completeness metric called, these categories called full splice match and incomplete splice match. We can see as we go from 300 base pair bin to the 500 base pair bin, there's a sharp increase in percentage of reads that are full splice match. So that's a good quality metric for isoforms. And so, this gave us a lot of confidence that these read lengths could span multiple splice junctions and even resolve full length isoforms.

And this is just a read level example of just that. We have the CD1C gene and we see those reads spanning the multiple exons. The bottom two are known isoforms and the top one you can see that second to the last exon missing. It's not an annotated isoform. So that's pretty exciting.

And so I'd like to acknowledge the immensely talented scientists in my group. The PRISM team, Broad clinical labs, and our industry partners. The single cell experiment in particular was an intensive effort and it took everyone coming together to make it a success. And it's been a sincere pleasure collaborating with Roche on this transformational technology. Mark, congratulations on this landmark achievement.