I will present an overview of GSK's high-throughput expression platforms and the advantages of employing them, such as streamlining delivery of protein reagents, enabling multiparameter optimisation of complex reagents generation. In addition, they also enable collecting vast amounts of well-curated data for human and machine learning. This data was used to build and parameterise a model to predict protein expression in various systems.

Closed loop autonomous learning for protein engineering is a vision of a possible future that may connect AI to physical hardware, resulting in experimental design, execution, and analysis, while minimising human input. This is attractive, as it offers the possibility of more rapid discovery and development of therapeutic lead candidates. This talk will present the ongoing work at Novo Nordisk and our collaborators, and present some finding and future directions.

The SGC, a global public-private partnership, uncovers novel human biology through structural genomics and chemical biology approaches. Target 2035 aims to develop tool molecules for every human protein by creating massive open datasets of high-quality protein-small molecule binding data, using DNA-encoded libraries and affinity selection mass spectrometry platforms. Models built from these data will allow prediction of new and more drug-like small molecule binders, which will be tested experimentally.

Artificial intelligence and deep learning have significantly advanced structural biology, achieving unprecedented accuracy in modeling folds directly from amino acid sequences. Despite these advances, deviations from empirical structures are not uncommon, and experimental determination of protein folds remains vital for the advance of structural biology and biomedicine.

Thanks to progress in high-throughput DNA synthesis and sequencing, artificial Intelligence and machine learning have emerged as leading approaches for building sequence-to-expression models for strain optimisation. In this talk, I will discuss our recent progress on using this technology for designing novel regulatory and coding sequences with improved expression phenotypes, using a combination of supervised learning and optimisation algorithms.

Despite recent advances in our understanding of genetic features that promote robust protein expression, transgenes in biotechnology have remained largely unchanged for decades. Natural human genes are rich in intron sequences that can drive these crucial expression benefits, but were previously difficult to replicate in artificial transgenes. At ExpressionEdits, we're changing this by deciphering ‘genetic syntax’ using high-throughput screening and machine learning to design intronised transgenes with improved protein expression.

Manufacturing proteins in CHO and microbial cells faces challenges in maintaining high cellular productivity over many cell divisions. We have developed a plug-in gene technology that prevents cell line production instability. The plugins link cell growth to antibody secretion using biosensors that regulate essential genes in CHO cells and beyond. This technology enables stable production and supports continuous manufacturing, improving scalability and commercialisation of antibody therapies.

Discover how CSL Behring’s digital ecosystem accelerates scientific progress in protein production. This presentation offers an overview of integrated digital tools and platforms that streamline workflows, enhance collaboration, and improve data management. Through real-world examples, learn how optimised processes drive key performance indicators, enabling more efficient, scalable, and innovative R&D efforts that support breakthrough therapies and advance biopharmaceutical development.