Listen on
Overview
Many biotech and life sciences organizations struggle with fragmented data and siloed laboratory workflows, hindering regulatory compliance and slowing down critical drug development. This isn’t just about managing samples; it’s about integrating disparate instruments and systems to generate reliable, high-throughput data for critical decisions. In regulated environments like GMP/GLP, manual or piecemeal approaches introduce errors, slow down IND filings, and prevent scaling. Executives need to accelerate drug development, ensure compliance, and optimize R&D spend.
In this episode, host Ross Katz speaks with John Liberty, Senior Bioanalysis Technical Sales Consultant at Thermo Fisher Scientific, who shares his insights. With a background spanning GMP labs and startup assay development, John brings a practical perspective to implementing specialized LIMS solutions that directly address these industry challenges.
Our conversation covers the specifics of Watson LIMS for bioanalysis, its implementation in both early-stage startups and established laboratories, and Thermo Fisher’s broader vision for Connect Enterprise, a platform designed to unify the entire lab environment. We explore how these tools enable a shift from instrument-centric to data-centric operations, helping organizations manage complexity and drive innovation.
Key Takeaways
Fragmented lab data directly hinders regulatory approval and operational scale.
Manual data handling and siloed instrument outputs in bioanalytical laboratories introduce significant risk in regulated environments. Without a unified system like Watson LIMS, organizations face delays in IND filings and struggle to scale operations efficiently, impacting time-to-market for new therapies. This challenge extends beyond individual experiments to the entire R&D pipeline.
Implementing a specialized LIMS requires adapting business processes for long-term data integrity and efficiency.
Adopting a system like Watson LIMS goes beyond software installation; it demands internal process adjustments to conform to structured workflows. This upfront work ensures data consistency, automates compliance checks, and provides the foundation for scalable, validated bioanalytical operations. Companies must be prepared to meet the system halfway to realize its full benefits.
True lab automation requires a layered data integration strategy across instruments and enterprise systems.
Beyond direct instrument integration, managing the modern lab necessitates a platform approach. Tools like Integration Manager bridge data from diverse instruments, while broader platforms like Connect Enterprise unify scheduling, LIMS, and ELN systems, reducing application fatigue. This creates cross-functional data flows essential for increasingly autonomous operations.
Automating routine lab tasks frees scientific talent to focus on high-value innovation, not just compliance.
While regulatory compliance is a primary driver, the significant return on investment of systems like Watson LIMS and Connect Enterprise lies in their ability to automate repetitive, low-value tasks. This allows scientists to dedicate more time to complex method development, experimental design, and critical data interpretation, accelerating research and development pipelines and maximizing intellectual capital.
Related: CorrDyn helps biotech and life sciences organizations build data reliability into their systems, from data engineering to broader digital transformation initiatives and full systems integration.
Full Transcript
John Liberty: Thermo’s really trying to be proactive. That’s something even with Connect Enterprise — making sure that it is application agnostic and that we can actually prepare for the future in a way.
Jason: Welcome to Data in Biotech, a podcast from CorrDyn where we explore how companies leverage data to drive innovation in life sciences. Every two weeks we sit down with an expert from the world of biotechnology to understand how they use data science to solve technical challenges, streamline operations, and further innovate in their business. Here we go.
Ross Katz: John Liberty, welcome to the Data in Biotech podcast.
John Liberty: Thank you very much. Happy to be here.
Ross Katz: Awesome. Well just to kick us off, would you mind giving us an introduction to your background and what brings you here?
John Liberty: Absolutely. My name is John Liberty. I started my career at Charles River Laboratories. I was working in a GMP lab, primarily doing analytical services, helping with routine testing, whether that was something like plate-based assays and ELISA, or a lot of the compendial testing like osmolarity or endotoxin. Then I transitioned into a startup company where I was primarily responsible for assay development surrounding PK and biomarker work. I was also helping to outsource that, building relationships with different CROs as we continued to push our program forward. Following my startup experience, I moved into my current role at Thermo Fisher, where I’m part of DSAS, the Digital Science and Automation Solutions group. What we do is work with different groups and potential customers to help implement different LIMS solutions. Now we’re currently partnering with our lab automation team to help implement some of our lab automation solutions.
Ross Katz: Can you give us an overview of Thermo Fisher’s work and specifically what your department does inside of Thermo Fisher?
John Liberty: Absolutely. To be honest with you, I probably couldn’t even give a high-level overview of Thermo in general. Thermo is anywhere and everywhere, and anyone who works in this high-tech industry has heard of Thermo Fisher, whether it’s in one capacity or another — getting reagents, or now that we’ve acquired PPD and working with some of our bioassay services, or even beyond that. Thermo is pretty much everywhere. They have their hands in every sub of the process and can really help drive you from start to finish, which is really great. My group in particular focuses on digital solutions, helping with LIMS, or if you’re not familiar, a laboratory information management system. What that does is help you with sample tracking and sample inventory all the way up through data and assay management and then reporting. In particular, I specialize in something called Watson LIMS, which is a bioanalysis-specific LIMS solution.
Ross Katz: Looking at Watson LIMS with the context that it’s bioanalysis-specific, would you give us an overview of Watson and what makes Watson bioanalysis-specific?
John Liberty: Absolutely. It does have a lot of general capabilities that you would see in most LIMS. That can include things like sample inventory and sample management and real-time sample tracking. It also does things like automatic freeze-thaw tracking, which is really important especially as you move into a more regulated environment. But then specifically when it comes to assay management and reporting and data management, it is very specifically tailored to bio-A. What that means is you can actually facilitate and build your bioanalysis-specific protocols into the software, and then it will help support that process as you move through your workflow. It also integrates with a lot of bio-A specific instruments, whether that’s something like ELISA or HPLC or MSD. It has those integrations in a few different capacities. You can have what we call built-in digital interface, or we also utilize one of our other products called Integration Manager, which can help bridge the gap between any sort of instrument that doesn’t have a built-in integration to Watson. What Integration Manager’s responsible for is it actually takes the data output from your instrument and transforms it into a file that can be easily digested by Watson. Then you can actually import that data right into your study. On the reporting side, it has something called CDISC reporting as well as other out-of-the-box reports, which help drive the regulatory process. It’s all specifically tailored to meet bio-A specifications and guidelines as per different compliance bodies and things of that nature. It also drives our PK, which is pharmacokinetic module. If you have any PK modeling, which is pretty much always necessary when you’re doing drug development especially in the pre-clin or once you move into the clinical space, that’s something that is supported in Watson as well.
Ross Katz: Awesome. I want to dive into each segment of functionality there — the data integration, the PK modeling, all of that. But before we do, I’m interested in when is the phase of a company’s lifecycle when they decide that Watson LIMS is the right solution for them, or what are the challenges that they’re facing at the moment when they decide yes, this is a new tool that we need to bring in and implement?
John Liberty: That’s a great question. It’s really company-specific. We see companies that are 20 years down the line that are currently using paper-based solutions and really have come to the crossroads of we need something more modernized and more flexible. We see companies all the time that will come to us and say all right, how do we overhaul our current solution, these are our challenges, these are our bottlenecks, and then they want to implement something like Watson. On the other hand we have startup companies that are just starting to build out their program, just starting to build out their workflows, and they want to get this started from the jump. There are barriers depending on which road you take. As with any type of solution in this industry, it does come with a price tag. If you have a startup company that maybe is a little bit more limited on funds, that could not necessarily be realistic from the jump. We are working to do more startup-based modeling with our pricing, so that we can try to get that solution implemented from an early stage because Watson is built in a way that you can get it established and then continue to expand on it over time. It’s not a one-and-done type of transaction. Thermo really wants to partner with our customers to continue expanding things like the number of licenses that you have or the types of modules that we offer in our solution. We definitely see both sides of the coin. Either doing it right from the jump or doing it far down the line. If you’re working with a company who may be in a GMP or GLP environment and they’re looking to implement something like Watson down the line, it can take a little bit of time because, as anyone who worked in a regulated environment knows, there’s so many stock gaps and so many procedures and processes and change requests and things like that to follow before you can completely overhaul the way that you’re handling data and logging your samples and initiating your laboratory workflows.
Ross Katz: I’m interested in what that implementation process looks like if you’re already coming into a GMP or GLP environment, because you’re going from a world where all of these processes are, I’m imagining, triggered from external systems and then patched together through people processes, to integrating the data and triggering them all from that centralized interface. Am I thinking about that right, or how does that look?
John Liberty: No, that’s exactly right. Especially when you go into the current bio-A space, there’s a lot of fragmentation going on. You have all these different instruments that have their own instrument software, and they have different file outputs that may be a little bit unique. One of the greatest challenges that we see in the industry today is being able to really connect all of those different pieces. Because one lab may not be doing the same assay as another type of lab. If you want to have an off-the-shelf solution, which is what we consider Watson to be, it needs to be able to handle all of these different scenarios and all these different types of output. That’s definitely a challenge that we see today.
Ross Katz: Obviously there are lots of LIMS in the LIMS space. I’m interested in, even within Thermo Fisher as I understand it, there’s multiple LIMS systems available for organizations to purchase. What differentiates Watson LIMS from some of the other LIMS that are out there in the ecosystem and in your portfolio as being the LIMS that’s custom-tailored for bioanalysis?
John Liberty: Sure, absolutely. In general, having the Thermo Fisher Scientific name behind our products really delivers a type of trust in what we actually offer because Thermo has been a staple in so many laboratories for so long that it really helps establish our customer base and speak to the quality of products that we’re putting into the market. That’s a really great thing to have in our corner. Within our portfolio we have different solutions that are tailored to different industries. One of our other solutions called SampleManager is much more versatile. It has the ability to work in different industries such as oil and gas, food and bev, manufacturing, or even large pharma. That’s widely popular as well. Where we differentiate here is that Watson is specifically built for purpose for bio-A. Anything that’s not in the bioanalytical space would automatically move away from Watson. That again speaks to the types of assays that it supports, the way that it regresses data, the models that are already built into it, the reports that it generates with that data and so on. We do have other products that help satisfy those requirements in other industries. With any other product, there’s always going to be competitors and there are a lot of other really great solutions out there, but we really try to drive home the point that we’re taking a what we call one-Thermo approach. Gathering that trust from customers that we have the LIMS solution, we have the instruments, we have the software, we have the knowledge — really helping to distinguish ourselves from the competitors that way.
Ross Katz: That makes a lot of sense both from the customer perspective where you want to have as few trusted suppliers as possible for the different pieces of equipment and process tools and software systems that you’re bringing in, but also from an integration perspective — you’re able to quality control the whole thing from end to end when you own each aspect of the workflow that’s being done. I’m wondering if we can ground the workflows that are being done in Watson LIMS in either a real example or a canned example — for example from sample receipt to validated report, what are the things that people are doing inside of Watson LIMS that the system enables?
John Liberty: Absolutely. I can give you a pretty basic example of what that would look like, a day-in-the-life scenario. There are a few different ways that you could initiate the workflow, but one of the most common ways that people will do it is you’ll start by designing your study. That could be done a few different ways in and of itself. You have what you call a manual study design build, which is a form that’s built into Watson where you can manually plug in the different treatments that you’re doing, your matrix, how many subjects you have, your time points, things of that nature. Or we offer what we call ISP or Import Study Protocol. This would be a pre-configured Excel sheet with various different column headers or fields that you can drop into Watson, and then Watson will digest that and populate your study in the platform. Once you accept that study design, it’ll populate all of the samples that are generated as a result of that study. Then what you’d be doing is going into the sample handling module and checking in those samples. That also could be done a few different ways depending on if you’re generating these samples in-house or if you’re receiving a shipment from a CRO. You can log different shipments that may be expected, whether it’s incoming or outgoing. If it’s something that you’re generating in-house you would go in and it has the option to check, is the tube cracked or is the sample not frozen, what is the status of the actual sample that you may be looking at. You can go to those drop-down menus and check in either one by one or in a bulk scenario. That could also be done through something like a barcode reader. There’s a few different options here. It’s not really a one-size-fits-all. What we call Watson is that it is configurable but not customizable. A lot of the functionality is out-of-the-box available to you, but how you tailor it to meet your needs and meet your workflows is pretty specific. Once your samples are all checked in, that’s when you go in and create a master assay. This would be in the sense of if you already had something validated and you’re ready to do a routine analysis. You would go in to create a master assay, which is essentially your protocol for your method. You would specify okay, is this an ELISA, is this LCMS, is it an HPLC, or so on. You would specify that and then link any instruments that are integrated with the platform. That is when you would go in and set your parameters. You would set your different standards or QCs and what those levels and numbers of replicates are. You would specify your regression model and where you want your cutpoints to be, where you want it to throw different flags if you have percent CVs that fall out of spec as per maybe your internal SOPs. Once you save that method, you can create a local run assay, and that is where you essentially take the template that you just created and just add your samples to it. You would designate what samples belong in that method and then save that as what we call an analytical run. Once the analytical run is set up, that is when the scientist would go into the lab and actually perform that experiment. Once you have the output from the instrument in the lab, that’s when you’d be using something like Integration Manager or the built-in digital gateway to move that data file from your instrument into Watson. Once that data is in Watson and linked to your study, that’s when you’re going to be running your regressions. You have the option to go through and view any sort of graphs or tables that may be available out of the box. As a user you have the option to go in, do a quick sweep of your data if you want, and then maybe pass it off to QA. You can have Watson throw different flags and create very clear lines in the sand of where something may have fallen out of spec. Once that review is complete, you can either accept or reject your run. If you accept your run, that’s where you could go in and start populating those reports. Those reports can all be exported to Excel or Word or something else, or downloaded locally as a PDF, or however you needed to function. Then that’s pretty much it. Those samples will be updated in inventory in real time if you have aliquots that are delicate-to-those-assays, or so on. In the case of doing something like validation, you would create a master assay that is specified as a validation run. That would allow you to pull those validation parameters like linearity, specificity, repeatability on those reports at the end. If you had a validated master assay already then you could use that for routine testing. If that’s something that you’re looking to establish early on, you would make sure that it’s specified as a validation run. That is where all those parameters would come into play once you start pulling the data together.
Ross Katz: I can see how the system has templatized the entire process from end to end. It reminds me of ERP implementations that we’ve done in the past where you’re taking a system that’s designed to consume a company’s entire business process from end to end and make it conform to this outline, this data model, this way of collecting information and making sure that it integrates together. One of the challenges that we always see with implementing a system like that is that any company that’s adopting the system has to meet the system halfway, or maybe even more than halfway — has to make their business process conform to the system in order to get the benefits of the automation, the integration, the efficiencies, the repeatability. I’m interested from your perspective in how Thermo Fisher navigates that, and what the conversations with clients are like when trying to make the business processes conform.
John Liberty: Absolutely. As you mentioned, there’s a lot of input that’s really needed from the customer to figure out one, what solution is actually right for that customer. Because as you mentioned our portfolio does extend across multiple industries and a few different workflows there. Really identifying first and foremost, what solution are you targeting? Then we have what we call discovery conversations where we’re actually doing requirements gathering. Our rule of thumb in my role as what we call technical sales consultants is our biggest thing is nailing the discovery conversation. What that looks like is a lot of active listening. We’re maybe asking some questions to probe some information, but for the most much we’re really just letting the customer talk. It’s really just about listening to what is working for them, what is not working for them. What are their goals for today, tomorrow, five years from now? And how do we really get them there? Getting a lot of that information out of the gate helps us understand what we have to offer them. It also really helps us prioritize what we need to deliver to them in order for them to invest on our solution. As I’ve mentioned earlier, we’re not really looking to just sell something and then essentially drop off. We’re looking to really establish these partnerships so that we can continue to grow. All of our solutions in our portfolio are built in a way that can be expanded. We’ve even worked with earlier startup companies that maybe get absolute baseline Watson. Maybe you have two, three licenses at most. You don’t have any added modules. You just want to be able to log samples, create a concentration-based data regression, and maybe spit out some R&D level reporting. But then they’ll come back to us after several years and say, okay, I want to add IRM, which is our immune response module, an optional add-on that specializes in ADA and NAb assays. Once they really get to that step in the process, you can easily tack that on to your environment because the foundation of Watson is already there for you. The biggest thing is making sure that you hit discovery early and often and making sure that you’re not moving forward with a solution until you fully understand the requirements that are being laid out for you.
Ross Katz: That makes a lot of sense. Another thing that I’m imagining comes up that relates to what you were saying about integration earlier is that if not all of your systems are Thermo Fisher systems, or if there’s assays that you’re using that are off the beaten path, or using machinery that hasn’t been integrated with Watson yet — I’m interested in, for those companies that are coming with systems or assays that don’t directly integrate with Watson, how does that conversation go and how does it work?
John Liberty: Absolutely. That’s something that’s so relevant to today. Because science as we know has continued to evolve and it’s actually even accelerating in their involvement or evolvement. We definitely have those conversations where maybe some customers will come to us and say all right, this is what I’m looking to do, this is this new system and instrument that I have, what can you do for us, and something that we’ve never seen before. That’s going to happen, it’s inevitable. But what you do is you really need to understand what they’re looking to actually bring into the Watson platform. More often than not we’re able to work with them through our integration manager tool. Integration manager as I’ve mentioned is actually instrument agnostic. It’s really meant to be able to bring in data from almost anything because it functions on a customized basis. It functions in what we call transforms. Transforms are primarily built by our services team who are excellent. The transform is responsible for converting a data output into primarily an XML file that is then ingested by Watson. We could do it with something that we’ve never seen before. But there are some limitations. I wouldn’t be truthful if I didn’t say that there are some scenarios that we’ve come across that maybe okay we can’t integrate with that. Sometimes it just can’t be done. However, more often than not we’re able to pull what’s being needed into the platform and we have found a lot of workarounds around that. Watson as a product has been around for more than 20 years. It definitely does need to continue to evolve with the times. It’s really great that we’ve been able to leverage this Integration Manager tool because no product from 20 years ago would ever be able to have the forward thinking in the types of things that we’re doing today. It’s nice that we have that product to fall back on to help get us there and bridge that gap.
Ross Katz: You mentioned that there are situations where you can’t integrate, so the Integration Manager tool makes a lot of sense for systems that are able to produce data in specific output formats or in certain structures — the transforms just convert it into XML and then ingest in the system, and then you can work with it inside of the Watson LIMS. I’m interested in, what are the characteristics of the systems that make it difficult for Watson or the integration manager to work with?
John Liberty: A lot of it is really just about how the data is organized. If you have a different file output and it has certain headers or categories, that really needs to be able to map to something that can be recognized by Watson. That’s the biggest challenge, is really just mapping those categories of data into something that is comparable in Watson. Our services team spends a lot of time really figuring out and picking apart those different output files to make sure that if someone wants to get certain values in, that can actually resonate with something that Watson already has built-in functionality for.
Ross Katz: Right. It’s schema matching between what gets outputted from the system to what’s in the Watson system, and making sure that those mappings are both logical to Watson as a system and logical to the customer in terms of being able to interpret the data that’s being ingested and is able to go along for the ride of the entire thing. Exactly. Yep. That’s exactly right. That makes a lot of sense. So you’ve got all of this sample metadata, you’ve got all of your experimental metadata, you’ve got information about the assays that you’re using and the systems that you’re using. At this point you’ve got all of the metadata and then the data from the particular experimental runs, analytical runs that you mentioned coming in. You’ve mentioned a couple times the different types of modeling that bioanalytical organizations are doing inside of Watson. Would you mind talking a little bit about the types of modeling you see customers do and what the use case for those models are? I understand that people who are indoctrinated into the biotech ecosystem might already know what they are, but our podcast has people at varying degrees of exposure to biotech.
John Liberty: Absolutely. If you’re doing something a little bit more mainstream like maybe ELISA, you might be using something like a 4PL or 4PL quadratic curve to do regressions for concentration-based analysis. If you’re doing something like qPCR, that’s something that might more tailor towards a semi-log regression model. Those are just two examples of things that are pretty common in the…
Ross Katz: Can you talk about what are the questions that people are trying to answer with those, and how Watson answers those questions?
John Liberty: I can more so speak to the concentration-based one. What you’d be doing there is taking something like a raw absorbance value from your plate reader and putting it into Watson. What that would do is create a regression based off the model like the 4PL where you’re solving for unknown concentrations of your samples. You have your standard curve with your QCs and checks in place, and then you’ll be getting those concentration values for your unknown samples. You can use that to create those flags and make sure that if you’re doing something like a validated run, all of those checks and balances are being met by something like QCs and percent recoveries and CVs and things of that nature.
Ross Katz: So the plate has PDF, and you’re using those different concentrations to estimate how absorption occurs at different degrees of concentration. That sort of thing? Yep. Exactly. Even in the case of an ELISA you could have an eight-point curve where one could potentially be a concentration of zero. You’re graphing all of those standard points at known concentrations, and then it will interpolate from that curve an unknown absorbance value and spit out a concentration value. Downstream of that experiment, how is that going into decisions? What are the types of decisions that either the bioanalysis organization or the downstream customer of the bioanalysis organization, the CRO, is trying to make using the outputs of that?
John Liberty: Sure. I can give you one very concrete example, which would be something like if you’re doing a pre-clinical study and you’re trying to figure out your dosing for a potential new drug that you’re trialing. You’re not sure exactly what your ROA or your route of administration would be, you’re not sure how high to dose the drug in order to get the efficacy that you’re looking for. First and foremost you need to understand what concentration is actually making it to the target system. That’s where a lot of that comes into play. In particular, in my experience at the startup company, we were doing drug development for neurodegeneration. I was doing a lot of testing on different brain tissue and trying to understand what concentration of our drug actually made it to that designated part of the brain. Otherwise you may need to adjust your dosing, or you may need to adjust your route of administration, because you really need to make sure that your target drug is reaching the molecules that it needs to and connecting through the intended pathways to get the desired clinical outcome.
Ross Katz: That makes a lot of sense. Another thing I heard you talk about a little bit earlier was this idea of testing for immunogenicity or for anti-drug antibodies. My understanding is that Watson helps with modeling that as well, if I’m remembering correctly?
John Liberty: Yep, that’s absolutely correct. That is the added IRM or Immune Response Module that we have for our environment. As you mentioned that is specific to ADA or anti-drug antibody and then NAb or neutralizing antibody assays. What that does in a similar way to the baseline Watson, it helps really guide users through the process. Specifically with ADA, it functions on a three-tier methodology. It goes through a screening tier and a confirmation tier and then a titer tier. What that really means is first you would take a sample and screen and say, okay are there any drug antibodies present in this sample? Yes or no. If it’s yes, then you’d move to your confirmation assay, where you want to make sure that there’s no false positives from the screen. That could be from a variety of different reasons. But you want to make sure that you’re actually confirming that, and what they do is they actually immunodeplete those samples in a way that confirms if those antibodies are actually present or not. If you move on to the third tier, that would be the titer tier where you’re actually quantifying how many antibodies have formed as a result of your drug. That’s something really important because you want to make sure that if there are any potential negative effects as a result of those antibodies or if the antibodies are actually inhibiting any of the effects that you’re hoping to get from the drug.
Ross Katz: Right. My understanding is that, if I’m remembering correctly, there’s a cutpoint analysis of where the limit is to what you believe your patients can tolerate in terms of the amount of anti-drug antibodies or immunogenicity at different degrees of dosing and concentration. Am I thinking about that right?
John Liberty: Yep, that’s absolutely right. Cutpoints will be established to determine if something really does fall in a positive or negative realm.
Ross Katz: Right. You talked about the integration module, but I know that there’s this idea of Connect Enterprise, this unified platform vision for where Thermo Fisher is going. Can you maybe outline that for me?
John Liberty: Absolutely. There’s a lot to this. I’m going to try to break it down into steps that are a little bit more palatable here. Just to set the stage, Integration Manager is a Thermo product as I’ve mentioned that helps transform data, and it is specific to some of our current solutions like SampleManager — you can use it on occasion, or Watson, a lot of our customers for Watson use it. Connect Enterprise functions on something completely different. Connect Enterprise is a brand new up-and-coming platform that Thermo is working on that is essentially meant to create a single UI across all of your current applications and really just enhance your digital ecosystem. What’s really powerful about that is that it is application agnostic. It can function with Thermo Fisher applications or third-party applications and softwares alike. The goal here is to not really replace anything that you’re currently using, it’s really just meant to enhance your digital ecosystem. If you have something that you’re using maybe like a third-party scheduler tool to help your lab staff schedule when instruments are going to be in use, and then you have your LIMS, and then you have your ELN or your electronic lab notebook, and then you have your reporting tools, those may be all different solutions. We see all the time that customers are using a wide variety of different applications. We definitely know too, even though Thermo Fisher’s everywhere, there’s no one who’s using exclusively Thermo Fisher. We thought, okay how can we actually capitalize on that? That’s where we came up with Connect Enterprise, where we’re able to pull in the UI and the functionality of all these different solutions regardless of if they’re Thermo or not. It functions on primarily APIs, but then we also have what we call Connect Transfer. There’s a lot of connect here, so it can get a little confusing. But the way I describe Connect Transfer is it functions similarly to Integration Manager, but it is exclusively for Connect. Through API and through Connect Transfer, you can link all your applications to a single login on Connect Enterprise platform. You can create dashboards and reports that are potentially cross-application information. If you wanted to initiate a workflow where you pull in sample information and metadata from your LIMS, but you want to simultaneously create a new lab notebook entry and you want to leverage your scheduler tool, you can essentially log into Connect Enterprise once it’s set up and access all three of those platforms through the UI of Connect Enterprise, and then it will all sync in real time.
Ross Katz: Interesting. Okay, let me try to see if I can understand what’s going on because there’s multiple levels of data integration happening here. There’s the level of data integration where, when you’re doing a Watson LIMS implementation, you’re orchestrating the entire bioanalytical workflow from end to end within Watson LIMS, but then there’s still going to be software systems that are outside of the scope of that end-to-end workflow — you mentioned reporting systems or orchestration systems or something like that. Anything where information needs to move from one system to another, or there needs to be a higher level of either triggering of work or movement of data, and maybe you want a single sign-on or a single place where you can log into all of those systems. That’s the part that doesn’t necessarily make sense to me because my understanding is that Watson LIMS is doing this orchestration, but then there’s also this opportunity to do it in Connect Enterprise. Can you help distinguish when a system would be integrated with Watson or through Integration Manager and when Connect Enterprise would be the umbrella that it falls under?
John Liberty: Absolutely. Connect Enterprise is — someone used this analogy the other day and I actually really liked it, so I’m going to recycle it. Connect Enterprise can serve really as the mortar to the brick wall of your digital ecosystem. It really can be in between all your different systems. In a perfect world, if you have a solution like SampleManager LIMS, you can link your SampleManager LIMS to your Connect Enterprise platform, and then if you’re using something like Chromeleon or some other lab automation system — in Thermo we have another lab automation tool called Momentum — if you’re using that, what you can do is create those workflows in Connect Enterprise and then pull metadata in real time from SampleManager and then initiate your workflow in something like Momentum or whatever else you’re using. Maybe you’re running an LCMS assay on Chromeleon or what have you. It’ll be updating and pulling reports in real time and will actually send that data back out to their home applications. If you were to log into SampleManager on your own after you initiated something, it’ll be updated and populated in real time because Connect Enterprise will be pushing that data back into SampleManager, but it’ll also be available in Connect Enterprise. You don’t actually have to go back into SampleManager if you didn’t want to. You can pull that information in. Watson and SampleManager, as they stand today, are just LIMS systems. Whereas Connect is supposed to be everything. It can tailor your LIMS system and pull it in and connect to your electronic lab notebook or your lab scheduler or any other types of reporting tools that you may have.
Ross Katz: Let me say back to you what I’m understanding. They’re just LIMS systems, just Laboratory Information Management Systems, a place where you house data — where the outputs of all of these different workflows and the metadata from all of these different machines and process steps are coming together so that you can generate the reports that you need for regulatory compliance and whatnot. But if you want to take a step toward full-scale lab automation where you’re able to connect the dots between all sorts of different systems, including your LIMS systems, your ELN system, your lab automation system — if you want to connect those dots then you need something like Connect Enterprise.
John Liberty: Exactly. The larger picture here is that Connect Enterprise, the scope of that product goes so far beyond just a LIMS. That’s the biggest thing here — the LIMS has its own scope of what it’s meant to do and how it helps drive your workflow, but anything beyond what a LIMS can do is meant to be caught by Connect Enterprise and really just link everything together in one place. Something I’ve heard talked a lot about is application fatigue. It’s so real. It is something all the time. Even when we’re talking about Connect Enterprise with customers, the last thing people want to hear is, hey you should go get this new product, you should go get this additional platform and get all this training of something completely new. I completely understand. It is so true. As I’ve mentioned earlier on in the conversation, with the fragmentation, people are responsible for learning and exploring and training on a wide variety of different applications and softwares and the ins and outs of everything. Connect really is meant to be a solution to that because if you can bridge everything together into that one Connect Enterprise ecosystem, you won’t necessarily have to go manually into each individual application. That is the larger picture here.
Ross Katz: That makes sense. If I may editorialize for a moment, one of the things that I’ve heard is also that it’s not clear where the boundaries of a LIMS system or an ELN end and begin, or lab automation system. Each of these systems has its lane, but the lanes overlap with each other, and we’ve got these puzzle pieces that don’t always — it’s not always clear which piece you should use for which segment of what the organization is doing. When it comes to thinking about tying them all together across an industry where different companies are using systems for different overlapping purposes, it’s understandably hard to imagine how it all fits together. That’s how it looks from the outside — would be interested on your take on how it is on the inside.
John Liberty: Absolutely. That’s very true. Even if you were to pull some information from something like SampleManager or different competitors and things like that, you could potentially get different answers. People are continuously adding functionality and features to their product to try to expand into those other lanes that they initially intended to not be a part of. The scope of different products can definitely vary greatly. It really could be subject to the use cases of where one product starts and where one product ends.
Ross Katz: In an ideal world, you’re helping to architect a startup lab from scratch. Let’s say we’ve got a startup biotech from scratch, trying to establish itself with the right data and process and regulatory infrastructure. It’s a small biotech that’s expecting to raise funding, expecting to grow quickly. I would love to hear how would you guide an organization like that in the lifecycle of when they need which piece of software — we could use the systems that Thermo offers. When should they be thinking about tacking on each element of it, and then where should they be looking as they scale into the actual enterprise — they’ve gone public and they’re now ready to put on their big boy big girl pants and go be a full-fledged biotech.
John Liberty: Sure. In my opinion, one of the best first steps to take is really the lab notebook, whether that’s an electronic lab notebook or otherwise. That’s where everything starts. If you’re trying to do assay development from scratch and you’re really trying to get your experiments off the ground, that’s where you’re going to begin. Having an ELN that can easily organize your data and initiate different reviews and make sure that it’s stored in a way that is reachable and you can easily recall it and things of that nature. Once you start scaling up and you get more samples in your lab and you have potentially different partnerships or you’re working off-site or doing shipments to and from a CRO, that is where the LIMS comes into play. It can be really easy to lose track of samples when you don’t actually have a system in place. Any scientist will tell you, you go into a minus-80 freezer and you may have started with the idea of having all these designated racks for storage and labels and all that, but all it takes is a label to fall off a rack because it’s minus-80 degrees Celsius, and then you lose track of the sample. It can be really easy to lose that sense of organization, especially once all the frost starts coming into the freezer and you need to de-ice it and all that. That’s really where a LIMS comes into place — when you start scaling up the number of samples that you have and you need to start really organizing your data in a way that can be prepped for a regulatory body. Also once you start expanding head count, you really need to make sure that that data is available to all the people who need access to it. That’s where something like a LIMS could help facilitate that transition from person to person and also make sure that that data is readable and easily reviewable. Once you have all of those other applications in place, that is where Connect Enterprise would come into play. As it stands today, Connect Enterprise is not an out-of-the-box solution. I talk about Watson as an out-of-the-box solution, meaning that you install it, you get it up and running in your ecosystem, and that functionality that is built into the product is available to you from day one. Whereas Connect Enterprise is really meant to be a fully customized implementation. We really need to see what you’re working with for applications, how are you actually logging in samples, how are you keeping track of your laboratory entries, how are you actually handling data storage and data maintenance? Then what we do is build a solution to tailor that. That’s where I mentioned it’s kind of like the mortar and the brick — you have the bricks already in place and then Connect Enterprise helps fill all the gaps between that and hold it all together. Connect Enterprise would really be the last piece of the puzzle in those ecosystems. Once you really get up and running and you have all your applications working on a regular basis, that is where you can actually drive them into the future with Connect.
Ross Katz: That makes a lot of sense. It sounds like one of the reasons why Connect Enterprise seems so amorphous is because the mortar takes a different shape depending on which bricks you’re using. It sounds like it’s highly customizable and highly specific to the organization that it’s being implemented in.
John Liberty: Yep, that’s exactly right. We even have people going on-site to potential Connect Enterprise customers and doing workshops and POCs or proof of concepts. That is really where it comes into play — having something that’s already up and running and how can we make it better? How can we make it more easily able to communicate across the different applications? How can we continue expanding this so that, as science continues to accelerate, we can keep up with it and get ahead of it.
Ross Katz: I’m interested in how you demonstrate the ROI of Connect Enterprise to the companies that are adopting it. Obviously regulatory compliance is a clear win. It’s something that has to be done. But then there’s also the collaboration gains that you mentioned, and then there’s also the reliability, the repeatability, the reproducibility of the things that are happening. I’m interested in what are the ROI stories that you think resonate most with the customers that you’re talking to?
John Liberty: Sure. First of all those are two totally different conversations. An ROI on something like Watson would be completely different than the ROI in something or a company that would be looking to implement Connect Enterprise. With Watson, as you mentioned, regulatory compliance is a really big thing. In general we see a lot of customers who may be looking to move into the GLP space or actually maybe looking to file their IND papers and things like that. That is where you really need to start getting into a solution that is tailored to that and is really built to promote that compliance throughout. Whether that’s CFR, electronic signatures or FDA guidance and things like that. That would really be the biggest payoff if you’re moving from a non-regulated to a regulated space with Watson. With Connect, we’re actually working internally to figure out how to present that story in the best way. Something that’s really interesting is we’re working to build out what we call the new Customer Experience Center. That is a location we have where we have a variety of workflows initiated by our lab automation group and our Momentum software. We have it already integrated with our Connect Enterprise platform. We’re hoping to continue bringing perspective customers on-site so they can actually see it in action. What that does is really showcase the lab automation capabilities while also showing how Connect drives that process. I’ve been part of one conversation with perspective customers with this what we call the CXC, Customer Experience Center. What it does is you initiate your workflow that’s pre-configured in Connect. You click go, essentially. What it can do is pull that information from SampleManager for your sample metadata, then you’ll move into the wet lab and you’ll see our autonomous robots grabbing plates and aliquoting different samples and then running things like an IC50 assay or an LCMS on our Vanquish. Then it will use Connect Transfer as I’ve mentioned to pull that data back into Connect Enterprise where you can populate the reports. What you see is fully autonomous lab with minimal human intervention. What we really try to drive with that messaging is that if you have a workflow that’s already solidified, you can just implement the lab automation piece to essentially have that on loop. You can skyrocket your capabilities by having something run on loop overnight with minimal human intervention. The goal here is not to replace human intervention, it’s really just to change the focus on where the human intervention comes into play. If you have robots doing something like aliquoting or placing a plate on an IC50 plate reader, you can have your scientists focusing on other method development or continuing to drive your portfolio forward. It’s really just shifting your resources into where is the best use of your time, and how can we eliminate the need for manpower on low-value or highly repetitive tasks. That’s the ROI that we’re trying to paint with that CXC.
Ross Katz: The facility that you mentioned with a fully autonomous lab in the future, I would love to see that at some point. That sounds pretty amazing and sounds like a good way of showing, not telling, people what you’re driving toward, which seems like the only approach that’s possible when you’re talking about something as amorphous as Connect Enterprise. To summarize that, it feels like we’re moving from a world that’s more instrument-centric to one that’s more data-centric. Would you agree or disagree with that idea, and how would you characterize it?
John Liberty: That’s definitely accurate. Data is the front-and-center concern that a lot of people have of how are we maintaining data security and how are we actually tackling data storage and making sure that things like audit trails are being captured, everything that’s really needed in a regulated environment. Being able to create solutions like Connect that is centered around data transfer and data storage, and making sure that we can actually keep up with the current evolvement of science. There’s always new instruments and new softwares and data types that are being generated and starting to flood the industry and the markets. Making sure that we’re really just keeping up with the direction the data is going in. Once the data’s out there you need solutions that can actually tackle it and store it and manage it in a way that is going to be beneficial to the users.
Ross Katz: The ecosystem is so broad and evolving so quickly. What I hear you talking about with Integration Manager and with Connect Enterprise is this idea that Thermo Fisher is trying to create the abstractions, the primitives that allow you all to interface with the broader digital science ecosystem and bring all of this data together. I’m interested in, where do you think Thermo Fisher’s role is in that environment, and how do you continue to stay abreast of where customers are going in terms of the tools that they’re using and the data that they’re trying to bring into that system?
John Liberty: Thermo’s really trying to be proactive. That’s something even with Connect Enterprise — making sure that it is application agnostic and that we can actually prepare for the future in a way. We’re not going to be able to predict what is coming down the line in one, two, three, five years from now. Making sure that we can create solutions that are here for the long term for our customers. Thermo’s been having a really large push and creating something that can serve you today but then also serve you in 10 years from now.
Ross Katz: There is seemingly infinite space for new drugs or new therapies to come to market and help improve people’s lives, but making the economics of the entire ecosystem make more sense from a funding perspective has the potential to create even more opportunities for science and raise all boats. I don’t know if you have any thoughts on that.
John Liberty: Absolutely. Like you mentioned, scientists — if you asked any scientist off the street, more often than not they’re going to tell you that doing something like a repetitive task of pipetting or routine analysis for sample prep isn’t necessarily something that they want to be doing all the time. It can definitely be new and exciting especially as you’re doing something like assay development — you’re doing trial and error. That’s the basis of science in the scientific method. That’s what’s really exciting. Once you get something solidified and validated, if we can have things like lab automation tools that can do the sample prep or the early on low-hanging fruit or something like that, then you can actually have the scientist working on the more innovative components of it and really be on the back end of the things that really can bring value to the company and to the group.
Ross Katz: That makes a lot of sense. As we head toward the end, where can people go to learn more about you?
John Liberty: Absolutely. Feel free to add me on LinkedIn or shoot me a message on there. Again, my name is John Liberty. We have lots of information and materials available on our Thermo Fisher Scientific website. In particular if you’re interested in a LIMS like SampleManager or Watson, or interested in starting conversations about our new and upcoming Connect platform, feel free to shoot right to the Thermo Fisher Scientific website and you should be able to find what you’re looking for. If anyone wants to shoot me an email, feel free to connect. My email is [email protected].
Ross Katz: Awesome. Well John, it’s been a pleasure having you on the podcast. Really appreciate it, and let’s forward to connecting down the line.
John Liberty: Thank you very much. It’s been great to be here.
Jason: And that’s it for this episode of Data in Biotech. If you enjoyed the episode, please subscribe, rate, or leave a review in your podcast platform of choice. See you next time.
