“We’ll see a shift from exploring what exists, to creating what’s possible”
The VIB Tech Watch team on Trends in Technology for 2025
December 17, 2024
As we’re counting down to the new year, it’s time to reflect, take stock, and make plans. Which new and emerging tools are bound to take the research world by storm in the next year? What new developments are here to stay, and which ones will rapidly be taken over?
Who better to ask about what’s next in terms of life science technologies than VIB’s own Tech Watch team: Wai Long Tam, Sarah Geurs, Toon Swings, and Samantha Zaunz. As this A-team scouts new technologies all year round, we’ve asked them to compile their long list of tech trends poised to make a difference in 2025.
Synthetic biology: a slow revolution
Just last week, the entire Tech Watch team was at ‘Next-Generation Synthetic Biology’ in Ghent. The VIB Conference is already at its fifth edition, so you can hardly call Synthetic Biology a new or emerging field. Slowly but surely though, it’s inching closer to transforming the way we study biology.
“Companies like IDT and Twist are highly skilled at synthesizing and assembling short DNA fragments,” explains Toon Swings. “Meanwhile, newer technologies are making it possible to synthesize much longer DNA fragments. Such methods could eventually eliminate the need for traditional cloning and pave the way for creating plasmids, or even synthetic chromosomes.”
One example the team highlights is NeoChromosome, a US company building on the work of Jef Boeke and others in designing a fully synthetic yeast genome. NeoChromosome specializes in genome-scale biological engineering including the de novo design, synthesis, and delivery of large DNA molecules and even chromosomes. Many other start-ups are emerging in this space, opening doors to applications that were previously impossible.
Wai Long Tam is convinced that these incremental improvements will turn out to be transformative: “We’re now looking at synthetic proteins and synthetic DNA. With the vast amount of data we’ve collected over the years, the idea of building a completely artificial cell from the ground up is no longer science fiction.”
The pathway to these breakthroughs starts with simple setups, often rooted in microbiology, notes Tam: “Synthetic E. coli and synthetic Yeast 2.0 already exist, building upon an existing biological chassis. However, as highlighted by Petra Schwille at the Next-Generation Synthetic Biology conference, a de novo build—from scratch—is likely not far off, with the potential to extend to human cells in the future.”
Tam is convinced that this shift will allow for much more in silico experimentation, using data to create a sandbox for innovation. “Many startups are now emerging to drive the field forward,” Swings adds . “For the time being, however, there are some major hurdles when it comes to accessibility and affordability.”
Read more about the present and future of synthetic biologyDynamic single-cell readouts
In addition to static single-cell analysis (e.g., snapshot technologies like 10x Genomics), the field is expanding towards dynamic single-cell readout platforms. These systems allow for the functional interrogation of cells over time, providing richer datasets for understanding cell behavior.
Sarah Geurs: “Current single-cell approaches typically require you to lyse cells, extract RNA or DNA, and read out from there. We’ve started testing two systems that can compartmentalize single cells and provide functional readouts of individual cells and their interactions before proceeding to sequencing: a platform by Lightcast and the CellShepherd by Arralyze.”
Such platforms first compartmentalize individual cells—either in microwells or droplets—and allow for the addition of various components. A typical example is to load cells to explore cell-cell interactions, or add antibodies or other reagents. These systems also include integrated visualization, enabling real-time monitoring. The idea is that researchers can then select their cells of interest, culture them if required, and process them for further downstream analyses such as single-cell sequencing or mass spectrometry.
“The key advantage of these systems is their potential for high-throughput, reproducible analysis and their ability to retrieve cells for downstream analyses. There are a lot of exciting potential applications, like CAR-T, where selecting the most potent cells could directly affect patient outcomes, but of course it would also be a game changer for fundamental research. For now, the platforms focus mostly on immunology applications because of the more straightforward workflows, but the scalability, standardization, and specificity in readouts make them attractive for a wide range of applications.”
Though these technologies show great promise, they are still being refined, says Geurs. “While compartmentalization and monitoring are currently implemented, retrieval of single-cells for downstream sequencing purposes is more challenging. With further development, and in response to our feedback, we expect these platforms to evolve and eventually hold the potential to become valuable tools for manyVIB research groups.”
Spotlight on protein
Another trend the Tech Watch team expects to continue into 2025 is the increased prominence of protein-focused technologies. Swings: “While the team’s primary focus has traditionally been on omics and genetics, there’s a clear shift toward protein technologies, and we’re seeing an increasing number of innovations in the proteomics field."
Sarah Geurs: “Protein sequencing was already on everyone’s list for 2024, but we’re not there yet. In my view, it is one of the most important advancements we can expect in the new year.”
In collaboration with VIB’s Proteomics core, Quantum-Si’s Next-Generation Protein Sequencer™ technology came to VIB. Geurs: “With this platform, you can read out the amino acid sequence and align it to a reference to analyze your protein sample and detect changes in the proteome. Other platforms, many of them still in stealth mode, are working on other approaches for de novo protein sequencing.”
Some technologies rely on new chemistry, while others rely on pore-based approaches—nanopores, as well as new pores that could allow for reading post-translational modifications. “It’s fascinating that all of the companies we’re currently in contact with are each focusing on different aspects and problems related to proteomics and protein sequencing, making it particularly interesting for us to keep track of all of them and monitor their progress.”
Geurs is genuinely excited about the potential: “Just like we can now easily read DNA and RNA, it would be amazing to be able to sequence proteins directly. Each of these layers provide powerful insights, but to get a complete picture, you need to be able to analyze them all. That’s what makes protein sequencing such a thrilling prospect for the future.”
Next to sequencing, interaction studies are rapidly gaining ground, with companies like Depixus leading the way. These platforms enable the analysis of single-molecule interactions, from protein-protein interactions to DNA- or RNA-protein interactions, or other combinations, providing unprecedented resolution into biological activity.
All members of the Tech Watch team agree that such developments will open a whole new area of exploration. Swings: “There’s growing evidence that there is much more to discover on the level of DNA, RNA, proteins and metabolites than we previously understood. Only recently for example, technologies exploring the ‘dark proteome’ uncovered unknown mini-proteins linked to diseases.” (More reading tips from Swings on the central or cellular dogma here and here.)
Different protein sequencing companies aim to tackle the challenges at hand, claiming to unlock the vast space of hidden proteins. However, Wai notes, "if you don’t know what’s there, it is difficult to design tools for it. There are still many platforms yet to be developed to fully explore these areas."
With the arrival of Joana Pereira as a new PI at VIB.AI in 2025, Tech Watch is excited to team up with an additional scientific partner that has extensive expertise to push the boundaries and venture into unexplored areas of the protein universe.
Success story: Protein synthesis in 48h
The team was excited about the introduction of Nuclera’s eProtein Discovery, a device capable of producing proteins in just 48 hours. This tool, initially used for nuclear protein discovery but enabling in vitro testing, was tested in its alpha phase by Tech Watch and is now in an incubation phase and housed in the lab of Joleen Masschelein at the Center for Microbiology. The technology allows for the production of small quantities of protein—enough for mass spectrometry analysis—facilitating rapid DNA iteration and testing.
“It’s been a great example of the technology adoption process we pursue,” explains the team. “We tested it early on with two groups, the Masschelein lab and the team of Xavier Saelens at the Center for Medical Biotechnology. Now, after a call for proposals within VIB, 19 groups are actively using it—far more than we expected.”
This level of adoption highlights the broad interest in the platform as a fundamental component for research and as a way to test AI-driven hypotheses quickly.
AI everywhere—Integration across platforms
To state the obvious: AI is progressing at an extraordinary pace, and will continue to do so in the new year. Its integration into various technologies is becoming seamless. Adding an analytical AI layer to tools that already generate data is something that can now be implemented relatively quickly.
“Take flow cytometers with AI integration, for example,” says Samantha Zaunz. “These are essentially standard cytometers enhanced with powerful computational capabilities. We’re also seeing AI incorporated into spatial omics, particularly for tasks like segmentation, and imaging technologies, where AI has become a natural component of laboratory equipment.” Even in sequencing, AI is playing a crucial role—like revisiting sequences to unravel hidden insights, such as the ‘mystery genes’ that Peter Carmeliet and his team are exploring.
These are all areas where AI implementation is advancing rapidly, often without the need for heavy promotion. Expectations have shifted—what used to be marketed as an AI ‘feature’ has now become an expected, built-in part of the tool: “There was a time when every company representative we spoke to would highlight how they incorporated AI into their platform or device. Those days have come and gone. Now, it’s an open door—it’s simply part of the technology.”
This evolution demonstrates how deeply embedded AI has become in life sciences technologies, transforming tools and workflows into smarter, more efficient systems.
Tech Watch has explored several AI-enabled tools, including AI-based flow cytometry cell sorters like the one developed by the Japanese company ThinkCyte. Zaunz: “They use advanced optics, producing waveforms of each individual cell that contain detailed data about its morphology, regardless of whether it has a more commonly used fluorescent marker or not. AI can then classify these profiles into clusters, using tools like UMAPs [Uniform Manifold Approximation and Projection, a dimensionality reduction tool to represent many-dimensional data in a 2D space] or other algorithmic parameters for cell sorting.”
However, this approach comes with challenges: “If the AI identifies five subclusters when researchers expect only three, the results raise questions: What does this mean? Where do we start validating? Validation often requires sequencing, which can be costly and time-consuming. Such hurdles make it difficult to adopt these technologies, especially since they often function as a black box."
Tam: “AI is both a blessing and a curse. If it confirms your hypothesis, great. If not, it raises more questions, especially when researchers can’t immediately access the underlying logic.”
Companies like DeepCell have attempted to address this by creating tools that allow users to trace the AI’s classification back to specific, explainable features, a field known as morphometrics. For example, a cluster might be defined by granularity or nuclear aggregation. “With the data we tested at VIB, this approach proved more tangible, guiding us toward the right validation steps,” the team explains.
> DeepCell and ThinkCyte are both mentioned in The Scientist’s Top 10 innovations of 2024
3D Omics
Spatial and 3D omics technologies are rapidly becoming indispensable for mapping molecular interactions within cells and tissues. These platforms are advancing toward better integration with AI-driven analysis, enabling breakthroughs in imaging and molecular characterization.
Sarah Geurs: “The convergence of spatial and single-cell omics will drive clinical applications. Technologies like in situ protein and DNA analysis allow for more dynamic and precise studies, for example of the tumor microenvironment, or the interplay between neurons and glia.”
The Spatial Catalyst program, in collaboration with the Nucleomics Core, has already made strides in this area by installing advanced platforms, such as recently Element Biosciences’ AVITI 24. Swings: “This is essentially a sequencer that you can also use to analyze DNA and proteins in situ. You load your cells into the device’s flowcells and readout using the platform’s probes. While panel-based, these systems operate in suspension, adding a dynamic layer to spatial analyses. We’ll be further exploring these technologies in the coming year.”
Spatial omics is starting to follow a trajectory similar to single-cell sequencing, which has now become standard. Geurs: “It’s increasingly useful to include spatial context in analyses, as it allows us to examine the cell’s microenvironment and take all of its interactions into account. Instead of focusing on individual cells, spatial technologies enable us to study how cells behave within a tissue.”
That said, resolution still needs improvement. While subcellular resolution is often claimed, this is still a challenge. Significant advances are expected next year, with improved implementation likely on the horizon.
When it comes to going 3D with spatial technologies, the team is in contact with Stellaromics. “They assert to be the first company capable of delivering 3D spatial omics. We’re exploring a collaboration in which we can gather in-house data and count on the expertise of the Spatial Catalyst team to help us assess and validate it.”
Read more about the ins and outs of spatial omics at VIBEmbracing the black box with de novo approaches
Swings predicts that de novo technologies—approaches that start from scratch rather than relying on existing frameworks—will become increasingly prominent. “These technologies are critical for breaking out of the constraints of what’s already known,” he explains. However, as Tam immediately points out, these tools often function as a ‘black box,’ which can deter adoption: “Scientists need to understand the underlying biology, and when that’s not immediately clear, it creates a chicken-and-egg problem.”
The practical hurdles are significant. “Researchers often wonder how to publish results from these platforms, especially when the analytical data is proprietary, as is the case for the AI sorters we previously discussed. There’s a huge validation burden, which translates into significant lab work and strain on the researchers.”
Swings: “De novo approaches only seem to work well when they’re part of a dedicated project that accounts for the extensive validation required. Otherwise, people prefer to stick to what they know and expand from there.”
This reflects a broader tension in adopting disruptive technologies. While traditional research often focuses on verifying what we think is true, de novo approaches demand a willingness to step into the unknown. There is a need to ‘zoom out’, build a long list of possibilities, and then consolidate resources to validate the most promising candidates.
Swings mentions the fascinating talk by Chris Barnes at the Next Generation Synthetic Biology VIB Conference he just attended. “Barnes presented how he uses mathematical models to predict microbial communities in silico. Though this wasn’t really a black box type approach, it made clear once more how enormous the leap is from all the data you can generate to a validation in the lab.”
It’s another way in which Nuclera’s 48-h protein synthesizer (see inset) has been valuable. In their quest for bacterial pathways that can simplify peptide synthesis, the Masschelein lab was able to use the new platform to produce an enzyme that was difficult to produce through any other means. This allowed the team to accelerate the validation of their potential hit.
Ultimately, Zaunz observes, the two streams of research—hypothesis-driven validation and speculative exploration—must intersect and challenge one another. “A lot of research projects focus on stable pathways where you can validate pre-existing ideas. De novo projects, which rely on fishing for new discoveries, are rare and often limited to small-scale experiments.” However, when such exploratory projects succeed, they can uncover important breakthroughs.
Innovative & actionable
As we look ahead to 2025, the life sciences landscape is poised for transformative changes, with VIB’s Tech Watch positioned in the crow’s nest. “In the coming years, we’ll see a shift from exploring what exists to creating what’s possible. It’s an exciting time for life sciences, with technologies like AI and synthetic biology leading the way.”
While many new technologies are emerging, others are maturing and becoming standardized. Spatial omics, single-cell multi-omics, and image-based flow cytometry are all technologies that are evolving into standardized tools that will become more accessible for broader research applications.
Zaunz emphasizes the importance of high-throughput technologies in that adoption: “Technologies that enable fast, cost-effective high-throughput data generation are critical—whether in sequencing, phenotypic assays, or proteomics. These outputs feed AI models, driving even better results.”
As technologies advance, labs like the Aerts lab at VIB.AI are already harnessing these tools to unlock new biological insights, adds Zaunz. “By training deep learning models on large single-cell datasets, they can, for example, identify cell type-specific enhancer codes—shedding light on gene regulation and its evolution with unprecedented precision. This kind of innovation exemplifies how cutting-edge technologies and AI-driven approaches are working hand-in-hand to address complex biological questions.
As life sciences technology development accelerates, the Tech Watch team is focused on guiding researchers toward the most promising tools and platforms, operating at the interaction of innovation and practicality. Tam explains that their role is to act as ‘matchmakers’ between researchers and emerging technologies, connecting ideas with the right tools.
Geurs: “Many of the companies we work with are eager to learn from us, especially startups—to better understand what academia is looking for and what researchers need. Being able to provide this kind of insight, and to be part of that entire development process is incredibly rewarding.”
Tam: “We rely on input from both the research community and from core facilities to identify those needs and test technologies. By the time we hand over a validated tool, labs can integrate it seamlessly into their workflows.”
New, newer, newest: the different phases of technology de-risking
Tech Watch follows an entire pipeline of technology evaluation across different development stages. “Typically, we’re involved right from the beginning, in the early phases when there’s just a paper or a patent,” says Swings.
Geurs quips that sometimes the team is even too early: “When we reach out to companies that have just filed a patent for example, we sometimes hear, ‘Oh, it’s great that you’ve found us, but we’re not quite ready yet to go public’.”
When they are ready, there are three typical routes the Tech Watch team takes. Geurs: “On the one hand, you have demos: the device comes in briefly, researchers run their samples, and it goes back to the company. On the other hand, there are funded projects where VIB PIs apply to get access to a new device or piece of equipment for de-risking purposes.”
Once a technology or device has been commercialized, it becomes lower-risk. “At this point, we refer to it as adoption funding—we support technologies that show promise but expect them to work nine out of ten times.”
“From there, we often integrate the technology into a facility. VIB Technologies provides an excellent structure for this process.” Tech Watch frequently works together with core facilities, which have a strong interest in new technologies and bring in essential expertise that the Tech Watch team may lack.
Recently, the Tech Watch team has been rolling out a third route of technology adoption. “It’s what we call ‘proactive projects’,” explains Zaunz. “In these cases, we aim to lower the time investment and financial risks for scientists by evaluating early-stage and widely applicable technologies ourselves. In a second phase, researchers can bring in their research projects and samples.” The approach has been very well-received by the VIB community and the incentive for the companies involved is to receive detailed feedback about the performance of their device.
Geurs: “Though financial support plays a significant role in evaluating technologies, we don’t want to position ourselves as funders, but as technology enablers. That is why maintaining neutrality is very important to us. We’re not another company, nor are we a venture capital firm, and that distinction is valuable to our partners.”
For Tech Watch, this proactive approach is also exciting because it allows the team to assess the tools in detail. “We learn where a platform excels, and for which types of projects it's perhaps not ideally suited. In the end, this is our ultimate goal: to know which tools are valuable for which specific research needs. This knowledge is crucial to identify where we need to invest in new technologies in the future.”
Read more
