Scientific Resources

Trends from SLAS 2025

The Society of Laboratory Automation and Screening (SLAS) International Conference was held in San Diego last week, bringing together diverse research and technology organizations to showcase and learn about the latest innovations in automation, high-throughput screening, and data analysis solutions.

With more than 200 scientific presentations and over 400 exhibitors, the SLAS2025 conference covered a wide range of topics including advanced therapeutic modalities, diagnostic tools, AI/ML solutions, and 3D cellular models.

Here, we highlight two key trends that echoed throughout the conference with a particular affinity towards the importance of high-quality data, tools, and reagents.

Trend 1: AI/ML models are only as good as the data that feeds them

Not surprisingly, a common theme focused on the adoption of generative artificial intelligence (AI) and machine learning (ML) to accelerate drug discovery. An area of continued interest is implementing AI/ML models to improve drug discovery workflows, particularly in the advancement of lead candidates following a screening campaign. While the models and output attract significant attention, the most important aspect is the generation of high-quality, robust data to feed these models and established assays to validate results. For example, following the early success of GLP-1 agonists, the idea of utilizing peptide ligands as therapeutics is experiencing a period of renewed interest. AI/ML models can be used to design peptide binders optimized for potency, selectivity, and favorable physiochemical properties. However, these models are often built from on datasets from hundreds of thousands of peptide interactions. That number may seem large to some audiences, yet it can often present limitations that impact the success of the model. Adopting technologies such as phage display, where tens or hundreds of billions of sequences are screened can not only rapidly identify novel ligands, but it also provides a massive high-quality dataset that can be used to build new predictive models that successfully achieve project specific goals. As the field continues to generate large datasets from screens, affinity applications, and others, how that dataset can improve the quality or accuracy of an AI model should be considered.

Trend 2: Sustainable solutions for drug discovery and beyond 

The SLAS community continues to support sustainable solutions, including the SLAS organization in its presentation of the conference to the researchers, engineers, and exhibitors that offer sustainable products and services. This year, we saw novel biodegradable consumables and an emphasis to shift from animal testing towards 3D cellular models for pre-clinical assessment of drug candidates and diagnostic applications. Interestingly, many of the diagnostic applications rely on affinity reagents that are developed through animal immunization. The field therefore needs more sustainable technologies to reveal high-quality affinity reagents for a variety of applications. Tango Biosciences’ extensive phage display libraries for antibodies, antibody mimics (monobodies, nanobodies, peptides, and more) are well suited for revealing high-quality affinity reagents for virtually any target. Our proprietary Avidimer platform is designed to reveal antibody pairs that bind to a single target (or complex) to efficiently translate towards diagnostic formats such as sandwich assays and high affinity applications. Importantly, the recombinant approach is amenable to protein engineering to tailor the affinity and selectivity, while also offering a sustainable solution to benefit patients and the environment.