0:01
So with that jumping right into it, I think we all know that biological systems are intrinsically and intricately spatially organised.
0:15
We've done a whole lot of the left side over the last 20 years between bulk and then last 10 years with single cell.
0:22
We've done a lot of reduction, organising the cell types, telling us where the parts aren't.
0:29
But until recently, it's been difficult to say where the parts have been in the tissue.
0:35
And then obviously microscopy has gone back 150 years.
0:39
So we've been looking at cells for a long time, but not with this high dimensional measurements that we can do now.
0:48
What's amazing about our system is that it really retains that single cell resolution while giving you high level measurement.
0:57
So the era of spatial biology is firmly here.
1:01
Methods started coming out in the mid 2010s by 2019 it was the method of the year, continued to be method of the year in 2020 and now spatial proteomics is coming up in 2024.
1:14
So there's a huge amount of excitement around spatial methods.
1:19
Hope you all are excited and I'm certainly really thrilled to be in this genre.
1:24
What's really awesome about this Vizgen is we recently partnered with both of you and throughout our product line its mostly centred over here, in foundational discovery and some preclinical work.
1:41
That's where our chemistry MERFISH 2.0 really shines and our instrument the MERSCOPE and MERSCOPE Ultra really shine.
1:48
But we've merged with Ultivue recently in the last six months and they bring to our table the ISP InSituPlex, which is a midplex protein solution going up to 12 markers in STARVUE, which is the analysis platform that powers the data insights that are coming off.
2:06
This is driving us deeper and deeper into the clinical market, which is a great space for us and for the fields to really be delivering on the promises that we're making to the community that we would really be impacting patients at the end of the day that are more efficient there.
2:24
So the merger's been a huge boost for our business.
2:28
We've got three platform technologies that one of which I'll highlight today.
2:33
But I talked briefly about the InSituPlex at STARVUE and mostly focusing today on MERFISH 2.0 and MERSCOPE Ultra.
2:43
So the MERSCOPE has got this high-powered high Multiplex MERFISH technology.
2:49
We get really high resolution.
2:51
We're working with 60X or 40X objectives.
2:54
So we have sub cellular resolution.
2:56
The dots are incredibly small and we're able to pick them up with high sensitivity.
3:01
We've been scaling this up to 1000 genes with fully custom panels.
3:05
We have some predesigned panels as well that are available off the shelf with quick and turnarounds, which is great.
3:10
And we have full species versatility because of best customizability.
3:14
So if you want to study a plant or a fungi, we're your people.
3:20
We can generate a panel and work with you to make sure that you're generating high quality to them.
3:26
And then by mid last year, we released the MERSCOPE Ultra, which has all the benefits of the MERSCOPE platform, but now has two flow chambers.
3:34
So you can sort of do your early screening blocks and samples with a smaller flow chamber, cheaper reagents and then move up to this bigger expanded 3 centimetre squared area.
3:47
And the other thing that's amazing is that it has really fast speed when it comes to these measurements.
3:53
So we can view 3 centimetres squared three times in one, which a lot of the other platforms can’t. When it comes with a data processing unit that's on board or below.
4:08
They can generate all of these files loaded.
4:10
You don't have to transport them up to the cloud and deal with any sort of relative regulations that your institutions may be applying.
4:20
MERSCOPE brings together basically smFISH and our proprietary barcoding system and that delivers on the MERFISH promise to give us these filing instruments.
4:33
Combined with MERSCOPE, it's a really easy to use, fully integrated platform.
4:37
A quick primer on how MERFISH works.
4:40
In case we don't know, these are sort of target RNAs in our cell.
4:52
So in image one, these 4 RNAs are lit up and you see there's one there on the sheet.
5:03
In image 2, we bring in new fluorophores, different RNAs light up and then we sort of do this round by round with the 1,000 gene imaging, we're going up to 10 rounds of three colours, so effectively 30 colours.
5:26
I have to keep my eyes in the slides and then we do barcode decoding that enables us to say, OK, this blue transcript is in the top left.
5:36
This yellow transcript B is in the middle and at the bottom of the cell.
5:42
And that's all done streamlined on the platform.
5:46
No interaction from you.
5:48
I really wanted to show you this slide.
5:52
This is a human ovarian cancer, 500 gene panel, 174 million transcripts were detected here.
6:00
This is actually with our alpha system.
6:02
So we played a few more tricks on there, but that video that's playing on the right is actually straight off of our visualizer that's commercially available.
6:11
Anybody can use it and plug their data in and really showcases how the data goes from this photo at tissue level takes picture view, where the cell types are, who might be interacting with whom, all the way down to the sub cellular resolution where you can say, OK, this transcript was closer to the nucleus, but this transcript is closer to the cell boundary.
6:30
You start to do those types of intracellular analyses.
6:36
Obviously there are a lot of platforms out there.
6:38
This is our preferred benchmarking study, of course, because it highlights us in a good light.
6:43
But I think what's really crucial here that they both did really well is they normalised these data sets as well as and really, I think one of the major drivers of that is the re-segmentation that was performed put all of these methods on the same playing field.
7:01
Because without that re-normalization, that re-standardisation, you're basically comparing different datasets entirely.
7:07
And it's hard to know sort of realise any sort of understanding in the middle.
7:14
They showed that we actually have the highest sensitivity of all these platforms.
7:19
And that's actually MERSCOPE on the top, which is compared with MERFISH, that blue line in the middle.
7:25
So we've actually made substantial improvements at the company level and it has excellent specificity, which is great.
7:35
You know, specificity, sensitivity, there's always a trade off, but we sort of hit that sweet spot incredibly sensitive while retaining that specificity.
7:47
The sensitivity is incredibly important.
7:50
When you talk about transcripts per cell, they seem a little lower than single cell numbers you're used to sequencing that could be up in the thousands.
7:59
Spatial transcriptomics generally were lower in the hundreds, mid hundreds.
8:04
So what we've done here is we've actually taken the data set and downsampled it basically approximating lower quality data, lower sensitivity.
8:12
This is at 100%, no downsampling.
8:18
Then we do 50% downsampling, strip out half the transcripts.
8:22
You see our transcript per cell drops approximately by 20% and 10%.
8:27
These are the UMAPs that come out as you decrease your data quality.
8:33
Basically all of your cell types falling in the middle, they stop being distinguishable.
8:39
The problem with lower quality data is that, you know, all of a sudden your B cells are being completely overlapped with your T Regs in terms of the clustering space.
8:49
Even though we know that these cells are distinct and we can highlight that these transcript cumulus over the tissue is obviously decreased.
9:00
That's why the transcripts per cell and the sensitivity is incredibly important, one of the challenges with MERFISH 1 is getting that high sensitivity and degraded samples.
9:14
So basically all samples have some other degradation.
9:17
It's employable.
9:19
A lot of this happens before you even get the block in your lab.
9:22
It's a surgical resection that happened in the operating room.
9:26
You weren't there because you weren't scrubbed in.
9:28
You're not an MD.
9:30
You have no control over that.
9:31
You can be as nice to your surgeons as you can, but these are the realities and difficulties of working in the clinical space.
9:40
So with high quality RNA, we tie all these probes all along these transcripts.
9:44
We get all of our readout probes down, we get very bright spots.
9:48
But when we have lower quality RNA, there are fragmentations, there are breaks.
9:53
Some of these probes won't mind either because their RNA isn't there or if there's a break point across that all of a sudden our signal intensity is going down.
10:02
So we want to be able to re reacquire those samples or reacquire those transcripts.
10:07
So we've brought in MERFISH 2.0 where we pulled it in.
10:12
A lot of the benefits that we brought in with our FFPE protocol, including our RNA anchoring and then added on 2 huge chemistry benefits.
10:19
So we've optimised the encoding probe design, the dilution buffer there.
10:23
So now we have more efficient occupancy at every probe site.
10:28
And on top of that, we've augmented the readout probe structure.
10:31
So in addition to having higher encoding probe occupancy, we have higher readout probe occupancy and that's driving with this higher brightness and higher intensity translates to more transcripts of the thing.
10:45
Now more transcripts actually leads to more cells because the first step in any analysis platform is filtering out all of the bottom feeding cells that don't have any real content.
10:58
So we with MERFISH 2.0 get a huge high dynamic range of cell type discovery because of this imaging platform.
11:05
We can go all the way from incredibly abundant cell types that are I mean 50% of your sample all the way down to cell types that are small fractions of your sample.
11:15
Because of this incredibly high resolution, we can do sub-cellular transcript blood localisation, which gives us this really high resolution cellular phenotyping.
11:25
We can do all the mapping of cell Atlas and you want to understand where tissue, where cell types are across the tissue and the different microenvironments that may exist within that larger section.
11:38
The signal to noise is a huge benefit coming out of this platform.
11:42
So we're getting clear data which enables deeper insights and all of this, I'd say where we're moving, continuing to move is looking at cell interactions and signalling pathways.
11:53
And this platform really provides the data density that's required to power those statistical methods.
12:02
So just visually, we look at MERFISH 1 on the top, MERFISH 2 on the bottom.
12:07
You can see we're getting a dramatic increase in the transcript counts.
12:11
This is a human brain sample.
12:13
The other thing that I do want to highlight is that we are getting sub cellular localization.
12:17
So that's DAPI, but the white is DAPI and you can see where these transcripts are falling.
12:29
Jumping back into more diverse tissue types. So this is a cross brain, breast cancer, lung cancer, breast cancer.
12:34
So this is actually a larger section and these are five different cores of DNA and then we're getting a strong increase in sensitivity from between 2 to 8 fold.
12:44
That'll really vary based on how high quality your tissue is to start with.
12:49
If it's high-quality tissue, you're getting most of those transcripts already.
12:53
We expect a low sensitivity increase with MERFISH 2.0.
12:57
But as the data quality or the sample quality is reduced, the benefits only increase with MERFISH 2, which is super exciting.
13:05
Beyond that, it's a highly reproducible method.
13:08
So this is 2 replicates of MERFISH 2.
13:10
You can see they're extremely correlated at one.
13:14
This is transcription on some MERFISH 2 versus 1.
13:18
They're also really correlated.
13:20
But you'll see that these are all above circle Y = X line, indicating that we're getting more transcripts of MERFISH 2 here.
13:28
These are correlation blocks.
13:30
This is MERFISH 1versus MERFISH 2.
13:34
This is our RNA seq.
13:36
It's exceptionally well correlated across the board.
13:39
So there's high data confidence which is good. Looking at human breast cancer a little bit deeper, these are five different pores from that TMA in the blue we see the MERFISH 1 data.
13:57
Pink is the MERFISH 2 data, pretty substantial increases again representing that between 2 and 8 fold increase where this really starts to yield benefits for you all and for us is that we're getting less cells that are dropping out through that filtering like I was just saying.
14:15
So this is on the top MERFISH 1 across all these five pores, we've got about 365,000 cells.
14:22
But if we look at the MERFISH 2 data, we get 650,000 cells and that's all driven by the transcripts that we're getting of cells.
14:32
On the bottom I'd just have a short disclaimer on how we're driving quality.
14:36
So low quality is roughly below 50 counts per cell up to high quality.
14:45
So if we look at MERFISH 1, MERFISH 2 data, they cluster well together.
14:50
So preprocessing is required to bring the datasets together.
14:55
But we can see that the MERFISH 1 is completely missing some of these cell types.
14:59
So we're highlighting the cycling myeloid cells here, but there are other cell types in the middle that are also missing.
15:05
So we're with MERFISH 2.0, we're really driving this increased cell detection and increased cell typing, which is great, especially looking at immune cells.
15:15
These cell types have a lot of RNA, innately in them.
15:19
So depending on the sample protocol, if a lot of those cells intend to preferentially drop out because those RNAse get activated and start eating up those cells specific RNA.
15:32
But with MERFISH 2.0 we're capturing those cells. Getting a little bit deeper, this is getting into the cell interaction piece of things.
15:43
So and then MERFISH 1 this is the cells for overlay from this core MERFISH 2.0 data.
15:49
And this is a spatial enrichment of the different cell types.
15:53
So here this is Ts and Bs and this is PVLs and endothelial cells.
15:58
And what you can see is if we look at the yellow one single square cells, MERFISH 1 data, we can see that they're well correlated in terms of spatial enrichment.
16:11
And in this and in the MERFISH 2 data, incredibly also we're seeing that same data, but what's different is the statistical power here.
16:20
So this is higher indicating that we're driving maybe correct insight, similar story over here.
16:33
This heat is up in the high hundreds compared to the low teens in the supply.
16:40
So it's really driving better insights here.
16:46
We're looking at human brain.
16:47
This is actually a customer data set.
16:50
So we're looking at a fourfold increase between the MERFISH 1 and MERFISH 2 data or six fold data plan.
16:57
So really driving a lot of customer satisfaction with this product.
17:01
We're super exciting to release.
17:04
Moving on to colon cancer samples, we're seeing a similar bump in sensitivity here.
17:10
And this is with Tubingen out in Germany.
17:13
They were super excited because they had to move forward and publish it soon.
17:18
And then one last one.
17:24
So a lot of groups, some difficulty to start out, but with a little bit of optimization and some extra sound and care, we're able to get a tonne that will cover the cells.
17:38
But with that, just bringing us back to the overview that Vizgen brings to the table.
17:43
So we have our protein InSituPlex assay, we have our analysis platform STARVUE, which is going to start to merge in with the RNA transcript world pretty soon.
17:56
And MERFISH 2 and MERSCOPE Ultra really delivering on the transcript side of things.
18:03
And that you know, I'd like to say thank you, VIZGEN and Ultivue merged last October.
18:11
Our motto is stronger together now, really bringing the proteomic side of the business into the transcriptomic side of the business and delivering all those benefits to you.
18:20
So thank you all for your time today.