Structuring Flow Cytometry Data Across the Entire R&D Pipeline

Welcome everybody to the webinar today. We're gonna be talking about how we can structure flow cytometry data for an AI ready future. And this is really gonna be a message of how using Luma and our flow cytometry tools together really does deliver on the dopmatics mantra of of better together across a lot of our products. I'm Phil Mounteney, in the group called Science and Technology. I'm working heavily in Luma. Now before we go and and talk about the specific FCS Express integration with Luma, I thought I'd explain just what Luma is. And if you look at these apps on the left hand side, these are actually all Dotmatics apps, and we'd include third parties in this use case as well. And this ecosystem looks a lot like a customer ecosystem where some of these are desktop tools, some of them are web based, some of them have databases. And before Luma, even at Dotmatics, none of them had a common data platform for them all to communicate with. So in my opinion, this data management and pros processing in Luma is really, really important. It allows all of these apps to push data into Luma via API, via files, via databases. Luma can then manage that data, put it into, a relational schema that's easy to understand. And then all of these apps can pull that data back out, but also pull data from the broader ecosystem. So it really allows great data flow, which, as you'll see, enables great workflow. We can include instrument data in here as well with our instrument and data ingestion of Luma Lab Connect. And then not really a focus for today, but we're doing material and ontology management in here, the first truly multimodal registration system. We have adaptive workflows for creating digital twins of wet lab workflows. You'll see the tailored user experiences or dashboards. And then the data management, which I talked about as being foundational, obviously, lends itself well to, artificial intelligence so that the data is well structured, well correlated before it goes in. Luma is an open platform as well. You can get the data back out by a rest, GraphQL, JDBC connections. And the system just allows us to to cover a lot of space. It allows us to cover chemistry and biology, material science, data science. Loom is a great data science platform. It helps your IT administrators. We can also cover a lot of scientific disciplines. So we have assay development, plus cytometry, obviously, characterization, sequencing, etcetera. And this breadth of, disciplines and techniques allows us to cover many different scientific modalities from large molecules to small molecules and everything in between. So now you've got a a flavor of what Luma does. Let me show you a dashboard that shows how Luma is able to harmonize data within the Dotmatics portfolio. So this is a, a dashboard within one of the Luma low code apps. And this one's particularly interesting because it allows us to bring data from Ohmic, one of our flow cytometry tools, and also from FCS Express and pull it all back into the same dashboard and treat it all in the same way. So this dataset now came from FCS Express. Some of the key features are that we can pull in all of the files. We can pull in, data from easy panel if you you've done design, ahead of your, your actual flow cytometry. And this is a plate based example in here where we can show the results and show the hits. But really importantly because Luma is a multimodal platform we can even link out to in this case antibody data since we were using the flow cytometry to add context and characterization to this flow cytometry data as well. So it's really important just to understand that Luma allows us this data flow, this harmonization, and as you'll see shortly, the workflow. So now I'll pass you over to Sean who can talk about some specific examples in flow cytometry. Alright. Thanks so much, Phil. And, again, my name is Sean Burke from the team at FCS Express. And, I'm sure as we go through the webinar today, you're gonna have a lot more questions, so make sure to enter, your questions in the q and a in the chat section. And I'm really pleased to come in and talk about, some of the, techniques and, tools that we have within FCS Express to support Dotmatics Luma. And before we kind of jump into that, I did wanna discuss why use FCS Express. I know there's a lot of different tools on the market. And when we work in FCS Express, the idea is, you know, you have all of these sorts of, different, tools and modalities that you might be using for flow cytometry. And FCS Express allows you to bring in things like cell cycle proliferation, the whole Microsoft Office suite of products and functionality, including PowerPoint and Excel like spreadsheets in the software. You even see GraphPad Prism in here. There's a lot of things that we can do to kind of support or further support, analysis in GraphPad Prism through FCS Express. Of course, we're gonna support all of the different sorts of, you know, conventional and spectral mass cytometers out there as well as supporting, high dimensional data reductions and, pipelines and tools like that. But the other kind of important part about FCS Express, and this is, applicable for, you know, folks wanting to use Luma as well, is that we have a research use only version of FCS Express. We have an IVD, clinical version of FCS Express. It's listed as a, medical device with the FDA. And between these two, we have something called validation ready. So if you need to meet any sort of twenty one part eleven compliance functionality with FCS Express, we can help you support that, more or less out of the box. Now when we talk about FCS Express, it's also good to get a, a little bit of a history or a background. Right? We've been, developing FCS Express since nineteen ninety eight. And I think some of the the key, kind of points in here leading up to this integration with Luma Lab Connect and Luma is, you know, in twenty twenty one, Dotmatics, completed an acquisition of De Novo Software, which brings in FCS Express seven into the portfolio of tools that are, you know, compatible, with, Dotmatics Luma. And you're gonna see later as well, like, really any tools we can bring in and use Dotmatics Luma. But here we are in twenty twenty five, you know, almost twenty seven years later after the development of FCS Express initially, where we have this platform where we can connect, to tools like Luma Lab Connect and Luma and really help, provide structure and quality and traceability for data to help you move you, into the future. Now, you know, generally, when I'm, in front of a live crowd, I I like people to raise their hands and, you know, tell me, you know, what are your biggest challenges with data and applications workflows. For those folks in the webinar, you know, I really encourage you to put in the chat. You know, let us know, what are the biggest challenges because, you know, what we kind of see as a way to guide you here is, you know, folks are working in data silos or having issues with data traceability, you know, results that come off of a a flow cytometer from QC. How do those, you know, match up and marry through to a final end result? They have issues with data searchability. The volumes of data and the size of data is just getting, kind of bigger and bigger. Even things like data quality and loss prevention and insurance compatibility. I mean, these are all known challenges in the field of flow cytometry, which, you know, a a pairing of FCS Express with a tool like Luma, really helps you overcome a lot of that. And to kind of guide the discussion today, you know, I I wanna start with, you know, what is the kind of big difference, right, when you're using other tools, that could be, you know, similar to LUMA in some way. But in a lot of cases, you're using workflows and data that are siloed. And you have to use workarounds, and you have to use manual data processes to get a result into a kind of structure and searchability and a a database that, can prepare you for an AI ready future. And when we work in dogmatics LUMA, that really enables that data to be ready for anything. And what you're gonna see today is that includes modeling of the data, includes charts, queries, and kind of getting you ready for generative and predictive AI, based off of the way that the data gets structured when we go into Luma. Now as a flow cytometrist, you know, how can Luma enable your results? And that's what we're gonna go through today in a few kind of specific examples where we look at FCS Express. We look at how that data gets sent to Luma, what it looks like there, some of the features and functionality, that we can help you support. And, again, there's a lot of, things that can be done with Numa. And what you're gonna see today are, you know, just the tip of the iceberg. You know, we really encourage you to reach out to our teams and see what this product and this product pairing can do for you. But to kind of guide you through, we're gonna go through quality control, traceability of instrument data and data quality between FCS Express and Luma. We're gonna go through, talk about cross domain integration. So, you know, talking about different modalities of experiments. Right? Flow cytometry touches almost all of science. And as you saw kind of what Phil showed, earlier, you know, having the ability to link up between something like antibody chemistry or chemistry experiments, directly back to the flow cytometry data is super powerful for your group as flow cytometrists or for the groups that are working upstream or downstream of that data. Also, working with, you know, kind of higher throughput data, walk working between applications like FCS Express and Ohmic and EasyPanel and Prism and, other tools that are out there, and the data processing involved is is very, very important for folks. So the first kind of example I wanted to start with is quality control and traceability and instrument data. You know, essentially, the current state for a lot of companies and businesses and biopharmas and, you know, even academic labs is that data from many cytometers, sites, and experiments, are not easily traceable back to instrument quality control runs. They're not easily traceable back to instrument information or metadata. And none of that is searchable or, say, AI ready, because researchers are just kind of ended up working in these siloed spaces. So because of that, right, you end up with, again, these kind of siloed spaces. You don't have the ability to say, you know, I did QC on the experiment this day. My end results came out the other day and kind of getting the idea of how that looks like. So what I'm gonna do is show you example in FCS Express of kind of how this functions and, you know, some features and functionality, in the software that are we think are very relevant for folks in flow cytometry and how that relates to Luma. So I did just open up FCS Express. And if you've ever worked in FCS Express before, you're you're probably familiar with this. It looks and feels and works just like Microsoft PowerPoint. We have objects on the page. We can move them. We can resize them. It's very easy to build these things up. And we work with tools like integrated spreadsheets that we have here that also help populate charts, and that provides you live updating as you move gates. And as information changes or data files change within FCS Express. And this is a quality control experiment. Right? We are looking at the the mean FITC or the MFI of, this particular dataset or datasets over a period of time. And what you would presumably use this for is that when you go back and you get final results from an experiment run, you'd be able to go back and say, hey. On day five that I ran this experiment or, you know, when I did the QC on, you know, in June at some point, was everything in spec? Right? And right now, that's a very manual and tedious process to kind of gather that information and share it across labs and teams. So in FCS Express, we now have the ability, you can see here, to export to Luma, right, and export, FCS files, export reportables, and other information. And, essentially, what we do here is just like any other kind of, FCS Express experiment, we click run. And when we click run, the data that's all, populated here gets, sent through LumaLab Connect into a Luma experience, which allows you then to, kind of store that information, search it, query it, and link it up between other experiments and experiment types and trace that information back to, where it came from. So with that, I wanna show you, I wanna be able to, flip this over to Phil here, and Phil is gonna come in and show you what this exact dataset looks like once it gets up to Luma. Thanks, Sean. So I'm gonna take you through the journey that the data takes after Sean has pressed the button to send to Luma. As he mentioned, FCS Express deals with files, and Luma Lab Connect is our file handling system. It's up on screen here. There's two modes in Lab Connect. We can either connect to an instrument so we could connect to your flow cytometer and centralize those files or to your other instruments. Second mode is that we can have software like FCS Express, like Prism, push its files programmatically via the API. Each row in this table is either an instrument or a folder waiting for data to be sent from software. So what happens is is that this these files get centralized so they're more findable and accessible, but we can make them more interoperable by parsing key parts of that data. So if I scroll down to my FCS Express results, this is where Sean is sending his data. I can see the folders. I can see, a process, and I can just pick one. What we actually operate on is the XML file that Sean sends from FCS Express. We create a parsed version of that file, our XML parser here, and we also package up the raw files, the FCS files that Sean has sent across. Now, come and speak to us if you want a more detailed explanation of how we operate on this JSON object. But that happens in Luma proper where we can design a, data model for that data to go in inside an app. And then within that app, we can build up what we call an experience or a dashboard based on that data. So here's the same QC run that Sean showed in his desktop software centralized in Luma so it's more easily shareable across your team. We have the metadata pivoted automatically for you. We have links out to all of the, attached, attachments and raw files. And then we have the main QC plots delivered in this nice interactive viewer. But, again, this is up on the cloud. It's central central. It's shared, and it's much easier for your team to collaborate on this now. So with that, I'll pass you back to Sean who's gonna show, a different example. Excellent. Thanks, Phil. And to kind of recap, you know, what we saw within FCS Express and Numa here as well is, again, we went from that current state of kind of disjointed instruments and labs and processes, and we've now centralized all of the information up to Luma. That includes the charts and the visualizations and supporting data, making that accessible to all researchers. And the experiment results are now easily traceable and searchable back to cytometers, to individual researchers, to individual quality control runs, and we can bidirectionally link. So if we have data, that was the final experiment run and it came from QC, we can click on that and go back and find the QC. We can look at a QC dataset and say, here's all the experiments that were run, based off of that. And, again, what this does is it helps us reduce complexity. It helps us bring down the time to find results and to get answers. It increases the data traceability and the data quality. And, again, this is making us better, what we do as scientists and better at working together as what we do as scientists. But I think start things are getting even more interesting when we start talking about cross domain data integrations. You know, working between different experiment modalities, enabling, again, more data searchability and data reporting. Because, again, flow cytometry is this key tool that's part of the wider discovery process. It involves lots of different techniques. Right? Some people are doing, sequencing. Some people are doing chemistry. Some people are doing antibody design. But, really, right now, the current state is that we don't have tools to easily link, search, and access results from different discovery modalities. That makes it also different to access and actions on key results from other modalities. So if you were to run a antibody titration or do, antibody kind of design experiment, you know, finding out the amount of antibody to use or the amount of antibody that was used to, you know, arrive at a specific ICEC fifty value is difficult to come at. And, again, you have limited data reporting and visualization tools when you work between these cross domain assays. So as another example, when we open up FCS Express here again, This is, an antibody titration experiment. It's part of a, antibody, kind of discovery process. And as we discussed, you know, we're in FCS Express here. We have the live, updating that, will link to our spreadsheets and our charts and also this really important key result. Right? This is the amount of antibody to use for this particular clone, as determined experimentally. And you can see that as I move a gate, that information updates in real time, which is really, really, helpful and powerful for FCS Express. But if I have multiple datasets that I'm working with, again, as I move between them, that result updates in real time, and I can, you know, again, move quickly and access all that information. But how do I enable this now to kind of work across modalities? Right? This flow cytometry analysis is just a, kind of smaller part of the experiment process. And, again, in my batch processing now, I have the ability to export information to Luma through Luma Lab Connect. So I'm going to choose the export information like this key result, information that's gonna be helping us, build up these curves in, Luma, information that's gonna support these experiments. And that way, people who are working as part of, you know, flow cytometry being another modality, within the wider discovery process are gonna be able to access those. So with that, I'm gonna flip over to Phil, and we're gonna take a look at how that information looks like when it gets up to Luma. Thanks, Sean. So I'll I'll spare you the Luma Lab Connect talk. It's the same data flow as we saw before. The file lands in Luma Lab Connect. Luma Lab Connect parses it, and Luma operates on it to build a data model inside this low code data app, and then you generate an experience or a dashboard on top of it. So here we say see the same data that Sean was just showing you in FCS Express. We have the two titration runs, and just like in FCS Express, we can switch between the two. And then in the bottom left corner, we have the key results. And, again, we always have pivots of the key data and links out to the raw files in these dashboards as well. But as Sean said, the ability to span across a multimodal, research effort is really, really useful. So we can use link outs like this key this clone keyword, again, as I showed you at the start to go and drill down into the provenance of that actual antibody that was tested. So this is from our registration system, from our protein design system, from our wet lab system, and even pulling in, additional characterization data from things like protein metrics or from PRISM as you'll see shortly. So back over to Sean where we talk about a a high throughput example. Excellent. Thanks, Phil. And, you know, before we go into our high throughput example, again, I just wanted to recap a little bit bit about, you know, what we saw. You know, previously in the current state, we have these, different sorts of modalities for antibody, design, antibody chemistry, things like sequencing or even graphing, that are all kind of disparate in a way. Right? And Luma is really going to enable this multimodal discovery by integrating across domains, you know, allowing you to search across not just flow cytometry data, but also kind of different, modalities that you might be working with. It provides access for all of the stakeholders in that system no matter, you know, where you're researching or what you're researching. And, again, having things in that structured format really prepares researchers for the AI ready future state. And, again, what Luma and, you know, tools like FCS Express and Luma are gonna help you do is reduce time. You know, time is generally money, so we're gonna help reduce cost. We're gonna reduce complexity because we're bringing information together into one location. And the other thing is opportunities for errors. Right? You've you never saw a copy and paste in all of this. We are sending information with a button click, you know, directly from FCS Express, through Luma Lab Connect up to Luma. And, again, this is going to increase data structure for that AI ready future state and also increase data quality. So we're gonna go into, the kind of last example that we wanted to walk you through today, as a use case for FCS Express and Luma. And this is around data processing, high throughput, and working across different applications, different software packages. And the current state is that researchers, you know, require lots of tools to process and distill results from high volumes of data. And, again, everybody knows that, you know, you have to use lots of different softwares and storage and application tool sets, that kinda help you get a result. But, that kinda help you get a result. But right now, it's not very easy to search or visualize through high volumes of data, metadata, and, you know, you know, key actionable results, especially across lots of different software platforms. It's also very difficult to access an action on a key result. I mean, if you have an IC fifty that came off of a flow cytometry experiment that's gonna support something else down the line, it's important to make that link so folks can do that. And, again, we have that limited connectivity between different software applications. And, again, what I'm kind of showing up on the screen here, you know, we have FCS Express. We have OMIQ, EasyPanel, Prism. There's lots of different tool sets, even ones that fall outside of the kind of Dotmatics, infrastructure. But we're bringing these together in a way through Luma and even through within the software applications itself that are very powerful. So in this particular, experiment that we're looking at, this is a ninety six well plate. And, essentially, what we're doing with tools like FCS Express is, kind of moving between different data files through our batch process. So we would kind of quickly and easily process through all ninety six data files here. We also like to look at this data in a plate based format because some days, we run one plate. Some days, we run ten plates. Right? We also have many facilities that run many hundreds of plates. Right? And that volume of information becomes very difficult, to kind of, pull together and process in a way that's, very, very useful. Now within FCS Express, we do have the ability to export to Prism. Right? So this batch action that we've had here for a Prism project for a while is is quite powerful. You know, when we look at results in Prism that come out of this experiment, essentially, this is what we get. When you run that batch process, you can generate all of the curves and visualizations that you'd like, directly by exporting from FCS Express to GraphPad Prism. But now we have the ability to not just, export the GraphPad Prism, but we can send any of that information to Luma through Luma Lab Connect. We can send any of the kind of, report informations, PDFs, PowerPoints, graphical data up there as well. Again, just through the click of a button, that will crunch through ninety six data files. It will crunch through three eighty four well plates. It will crunch through many, many different, sizes and formats that you want. And then we get that information up to Luma, which is gonna make it much easier to, work with and distill and share with others. So with that, I'm gonna switch it back to Phil so you can take a look at that application in Luma. Thanks, Sean. So here you can see this is that that same high throughput batch analysis represented in Luma. There's a lot of data in this one, a lot of raw files. So I've decided to to divide this up into three sections. The first section is links to the raw files and then that same pivot that we've seen in the other examples. Because this is a plate based experiment we can represent the plate heat map on the left or the hits on the right in exactly the same way that Sean had, decided to display in FCS Express. Finally because there was a prism analysis as a follow-up to to that we can also get this prism data in here. And I just wanted to take a moment to talk about how the data gets into this dashboard from prism. So if we open up that same Prism file in my desktop tool here, much like you've seen in FCS Express, and you'll see this in all Dotmatics software that is desktop, we have a send to Luma button. Remember, Luma Lab Connect is the thing that deals with files in Luma, so we actually send to Luma Lab Connect. We parse that data. So we parse the data tables out into, out of Prism, but we also regenerate the graphs. And it's that raw data flow that I've talked about before before that affords us the ability to pull in the data and the graphs into this dashboard as well. Great. Thanks again, Phil. And, again, to kind of recap what we're seeing here in this Dotmatics Luma state, we're pulling information together across many different softwares and applications. You know? In the the case that we just showed here, folks might be using FCS Express and Prism and even, you know, other tools like Excel and PowerPoint and PDFs. And Luma takes all of that information, puts it into one place, which allows you the ability to search and visualize the results across different software applications. And it does so in a way that helps you distill that information. So if you're working with low throughput to high throughput, you can do that in a repeatable and searchable fashion. And, again, we've been talking a lot about Dotmatics products. Of course, they're close to our heart. But you have to keep in mind that this is under the FAIR findable, accessible, interoperable, and reusable architecture where other applications can be linked up and brought into Luma. So it's not just Dotmatics applications. And, essentially, what we end up with as flow cytometry, if you're doing panel design, whether you're using FCS Express or OMIQ, if you're using Prism, Excel, other tools, we're pulling all that information into one place to allow you to access it and take action on it very easily. And, again, that reduces time. It reduces cost and complexity of experiments. Really, what we're doing here, again, is structuring data as well. We're linking up metadata, experiment dates, who ran things, what the results are. And that's the sort of information that we need to get ready for that AI ready future state where we can train models, based off of that data and based off of what we have up in Luma, to help give you more actionable insights. And, again, this is always helping us improve data quality. Information is being sent with a click of a button. It's being, structured into a way that's easy to distill and understand. So, again, I know, we've talked through a number of different kind of, examples today. I'm sure folks have a lot of great questions. I've been seeing some come through, in the q and a and the chat already. We're gonna come back through and answer those in just a few minutes. But keep in mind, Luma is the platform that's gonna fit the way that you do science. We know you lose lots of different softwares. We know you use lots of different modalities to figure out, you know, the really complex challenges and problems in science. And Luma is gonna enable your data to be ready for anything. You've seen the ability to work with modeling and charts and queries, kind of getting you ready, and get set up for generative and predictive AI based off of those results. And the way that we look at Luma is, it's this discovery amplifier. You're not just taking, one dataset from one software and one dataset from one software and equaling two datasets from two different softwares. Right? We're kind of multiplying that value. We're taking one plus one so it equals five. Right? And that's really what we wanna see our researchers getting out of tools like Dotmatics Luma. Also as they bring in, tools like FCS Express and anything within the Dotmatics Platform, Dotmatics tool sets or outside of it as well. So, really, again, our, our our kinda big goal at Dotmatics is enabling that lab of the future, and we wanna see you doing that today with Luma. And, again, we really appreciate you, spending some time, over the last half an hour to forty five minutes with us on this webinar to start discussing, you know, what's available, how it works at a very high level through FCS Express and Luma, and giving you some ideas about what you can do with tools like this going forward. So after the webinar, if you have questions, you know, make sure to contact us at support at denovosoftware.com. You know, we can make sure we answer any questions about FCS Express or get you routed to the correct channels with any questions about Luma. We really encourage you to visit the De Novo Software website as well as the, Dotmatics.com/luma website to start gathering some information. And with that, again, I just wanted to say, you know, a big thanks to everybody who's joined us on the webinar today. A big thanks to, Phil and the the team at Luma who's really, you know, made this possible, by working with FCS Express and working with our team and our customers to develop solutions, pass information to Luma and really help us out. So with that, we'll, take a pause, and we'll, take time to answer some of the questions that we've seen coming through. And, again, we really thank you for coming out today.