BOSTON—A record attendance of 7,549 scientists, pharma executives, and vendors—lots of vendors—left Boston this week feeling positive about the event and the state of the industry more broadly.
AI, organoids, and organs-on-chips were three of the prominent themes evident throughout the conference program and the trade show exhibition, as the Society of Lab Automation and Screening (SLAS) 2024 organizers had promised GEN in a preview last week.
Below we share some vignettes selected from a busy two-and-a-half days at SLAS 2024.
AI in abundance
One of the undoubted stand-out presentations at the conference came from Ava Amini, PhD, a senior researcher with Microsoft Research in Boston. “We haven’t realized the full potential of AI and machine learning,” Amini said. Until now, AI has been used largely for static tasks. The future of the field, she said, is to move to “active AI.”
In one vignette, Amini presented new algorithms that factor in an estimate of uncertainty in a given prediction, for example, the efficiency of small molecules to act as antibiotics. Moving forward, drug candidates could be ranked not only on their chemical potential but also the degree of confidence in those predictions.
In another exciting story, Amini presented an open-source framework called EvoDiff—evolutionary diffusion for protein design. As described in a 2023 bioRxiv preprint, the model has been trained on evolutionary-scale protein sequence datasets across the tree of life featuring some 50 million sequences. The model can then be used to generate new “biologically plausible” sequences that have not been observed before, learning to generate data from noise.
In a promising partnership with Boston-based Valo Health that began two years ago, Charles River Laboratories is promoting the “Logica Edge”—an AI-assisted approach to partnering with clients that allows them to “reach patients faster”—or fail faster, whatever the case may be. While Charles River’s forte may be in vivo mouse models, this platform offers access to thousands of computational predictive models, billions of virtual molecules, and Valo’s closed-loop discovery system, promising customers “a visionary risk-sharing model in a holistic suite of services.”
In the lead-up to SLAS, lab automation provider Opentrons launched two new products supporting drug discovery and microbiome research. The first is an automation marketplace that provides a one-stop shop for research tools, software, and services. Opentrons also launched a generative AI tool and protocol library to help researchers automate experimental workflows. The company is forming partnerships that can provide additional functionality to the marketplace, including companies like Genie Life Sciences, which provides lab orchestration software, as well as Cerillo, focused on automation solutions for microbiome research.
Opentrons’ new protocol library offers plug-and-play protocols for use with its robots. Within the library, researchers can find verified liquid-handling automation protocols for tasks such as nucleic acid extraction and NGS library prep developed by Opentrons as well as contributions from a global community of scientists.
The AI tool is akin to computational lab assistants like CoScientist, which was developed by researchers at Carnegie Mellon. The AI-powered protocol generation tool uses large language models to create novel workflows for Opentrons’ liquid-handling robots. It simplifies the process for scientists without the expertise to write their own code, those who want to focus on experiments rather than writing and debugging code, and scientists who only need to run the code for a short time.
The idea is “to lower the bar for what it takes to make our robot do something very complicated,” Gautam Prabhu, Opentrons’ CTO told GEN. “If you can get to the point where you are telling our robot what you want it to do in your own language, and it’s translating that into Python code, you have, first of all, saved a lot of time. You’ve also probably reduced the number of bugs and probably have a higher degree of confidence that it’s doing what you want it to do.”
The first AI protocols were for simple tasks like reagent transfer and PCR, but now it is developing automated workflows for nucleic acid extraction and next-generation sequencing library preparation. These workflows will require more complicated descriptive language and code.
Opentrons’ generative AI leverages ChatGPT but the company has made a number of customizations to work for its needs. The higher complexity of upcoming workflows “means higher chance of error. We have to make sure that we are keeping the accuracy the same because otherwise the usefulness goes down,” Prabhu noted.
Opentrons will launch an external beta for its generative AI tool later this year.
Organoids and organs-on-chips
Since the passing of the MOD 2.0 Act, interest in alternatives to animal models for drug discovery and clinical use has been growing. Alternative models like organoids and organs-on-chips featured in several presentations at this year’s meeting.
Anna Jezierski, PhD, at the National Research Council of Canada, presented on blood-brain-barrier-on-chip-models that were developed in partnership with SynVivo, a microfluidic organ-on-chip company, using human iPSC-derived brain endothelial cells that successfully replicates the behavior observed in vivo. The models are being used in preclinical assessments of CAR-T-based immunotherapies for glioblastoma among other studies.
Another presentation from Lena Smirnova, PhD, at Johns Hopkins University, featured brain organoids developed by combining iPSCs with gene-editing technology to study key neural development events and to assess the effects of different drugs and chemicals. The conference also featured presentations on organs-on-chips designed by Mimetas from company CTO Bas Trietsch, a gut-on-a-chip model from Alveolix that simulates peristalsis in vivo presented by co-CEO Nina Hobi, PhD, even a vagina-on-chip developed at the Wyss Institute for Biologically Inspired Engineering at Harvard University to study the microbiome of the vagina and test new therapeutics.
Corning, which develops both organoids and spheroids, also had a strong showing at SLAS including presentations on their 3D cell culture workflows and posters that demonstrate how to get the most out of their organoids. Elizabeth Abraham, PhD, market manager for Corning’s 3D cell culture business, told GEN that the company’s organoids and spheroids are finding use in a wide range of applications including studies of brain development, oncology, inherited diseases such as alpha-1 antitrypsin deficiency, infectious diseases including COVID-19 and Zika studies, as well as regenerative medicine.
Right now, much of the work using 3D cell cultures is preclinical but clinical use cases might not be far off. Organoids, for example, are attractive options not only because they derive from human cells but also because they could help lower drug discovery-related spending. As the SLAS presentations indicate, it’s likely that most organs can be successfully represented in these cultures. Some parts are easier than others in part because the research there is further along, Abraham noted. Intestinal organoids were the first to be developed and so scientists have had far more time to work on fleshing out the models. In contrast, prostate and lung organoids are a bit tougher.
And as 3D culture technologies continue to mature, automation will become increasingly important particularly to address scalability. For its part, Corning has made sure that its tools such as its spheroid plates are automation-friendly. Furthermore, Abraham noted that scientists have begun combining organoids with other technologies such as bioprinting and microfluidics platforms and CRISPR.
Swiss company Tecan laid out a suite of new product offerings in a wide range of areas including liquid biopsies, single-cell delivery, proteomics workflows, and lab management.
Launching in the first half of 2024, the Resolvex i300 is a positive pressure module for solid phase extraction and filtration that can be integrated into Tecan’s benchmark Fluet Automation Workstation, offering advantages for genomics and proteomics applications. Tecan also demonstrated the Uno, a compact single cell dispenser that enables single cells to be distributed into a 384-well plate in a matter of minutes.
The company also unveiled LabNavigator, an advance in lab digitalization and management. LabNavigator is a cloud-based digital product that builds an interconnected lab by “seamlessly combining lab equipment”—helping new staff quickly gain speed and confidence in running workflows, while improving the quality and reproducibility of lab pipelines.
From digitalization to quantum electrochemical spectroscopy (QES). Aiming to solve the “biology-AI gap”, Stanford University spin-out Probius has developed a small electrochemical sensing device that can rapidly measure thousands of analytes in a small sample by analyzing electron vibrational frequencies. The result is a multi-dimensional digital twin of the sample in the cloud, which can be interrogated.
Shuffling around the United Robotics Group booth, we ran into Kevin—a charming robotic device used for fully autonomous lab processes such as liquid handling, performing operations during nightshifts if need be. A detachable table provides human coworkers with a handy configuration tool. Kevin got his name, we are told, from the efficient independence demonstrated by the child hero of the film Home Alone.
GEN also met with Yokogawa to learn about the company’s new high-content analysis system. The CQ3000 is designed to capture high-definition 3D microscopic images of live cell cultures including organoid cultures over long periods. One of its key features is the proprietary water immersion lens mechanism, which supports detailed observation at high and low magnification levels. When paired with the company’s CellPathfinder image analysis software, scientists can get detailed visualizations of features like intracellular organelles and assess how cells respond to drug compounds.
This system will be launched commercially later this year. The company has other high-content analysis systems in this product suite that offer lower- and higher-end functionality than the CQ3000
Finally, SPT Labtech announced the release of firefly+, the most recent iteration of its liquid handling platform for NGS library preparation. This announcement follows one made at the 2023 American Society of Human Genetics, where the company shared several developments regarding firefly. This new release includes all the functionality of the original firefly platform and adds an on-deck thermocycler and increased labware capacity for those customers who want it. Like its predecessor, firefly+ works with various microplate formats and NGS sample kits and library preparation protocols including those offered by companies like ScaleBio.
Paul Lomax, head of genomics at SPT LabTech, told GEN that customers can expect the same reliability, ease of use, and reproducibility associated while enjoying more features. Scientists working with low-, medium-, or high-throughput experiments can set up multiple systems in labs around their campuses and facilities and pass along bespoke protocols and resources via a shared cloud environment. The applications team at SPT LabTech also develops and shares protocols in a community forum. Customers could come from academic core labs to large sequencing services providers to pharmaceutical and biotechnology companies.
SLAS heads west next year to San Diego, beginning on January 25, 2025.