Traditional biopharmaceutical production relies on one-size-fits-all expression vectors and iterative, empirical process optimization. This standard approach limits productivity and performance, hindering the commercial viability of increasingly complex modalities by causing development hurdles and extending time to clinic. In this talk, we present a vision for the future of biopharmaceutical development, and discuss our progress toward end-to-end design of genetic systems, cell lines, and bioprocesses using a portfolio of mechanistic, AI-based, and hybrid models. We highlight case studies that demonstrate the power of this approach across multiple modalities, including monoclonal antibodies and viral vectors.

As the demand for high-quality biologics accelerates, biopharmaceutical developers face increasing pressure to optimize cell line development for speed, quality, stability, and scalability. ATUM’s Leap-In Transposase technology transforms the protein expression landscape, enabling rapid, predictable, and high-yielding stable cell lines. Unlike traditional random integration methods, the Leap-In Transposase system facilitates highly efficient transgene integration at transcriptionally active loci, leading to enhanced expression, improved genetic stability, and accelerated development timelines. This presentation will showcase how Leap-In Transposase technology empowers biopharma teams to achieve faster workflows while eliminating the need for labor-intensive single-cell cloning. We will present case studies demonstrating its impact across monoclonal antibodies, multispecifics, and novel therapeutic proteins, highlighting significant improvements in titers, scalability, and manufacturability. Furthermore, we will discuss its seamless integration with CHO and HEK cell systems, making it a solution for early-stage discovery and commercial-scale production.

In the rapidly evolving landscape of cell line development and bioprocessing, powerful analytics with the integration of data models, artificial intelligence (AI) and machine learning (ML) as predictive tools offers transformative potential. To keep up with this expanding field, we introduce novel approaches for the development of cells and production processes, combining advanced analytical techniques and multivariate data analysis (MVDA). Applications range from manufacturing of classical protein-based therapeutics in CHO cells to gene and cell therapies, which represent promising avenues for treating a variety of genetic disorders and diseases. Implementation of these new tools for data analysis and modelling will further increase our process understanding and accelerate the pipeline to manufacturing. The workshop will interactively assess the audience’s use of analytics, data and modeling tools and then discuss use cases of powerful analytics and data-modeling for bioprocess optimization. Three case studies from our teams will show the potential of data- and model-based approaches: 1) Clone selection in CHO cell line development by combining MVDA with machine learning algorithms - Identification of clones with 40-60% higher titer in fed-batch 2) Predicting cell culture media quality and performance through untargeted mass spectrometry and MVDA - Fingerprinting approach that allows identification of bad media lots without in-process testing 3) Unraveling HEK-based AAV production through combination of UHPLC-Orbitrap-MS for the analysis of metabolome and secretome with targeted methods (e.g. trace element analysis and titer determination) - Generates models with >1000 metabolites to find markers for high AAV production