Big Data Opportunities in Biomanufacturing

Big data is not merely a storage requirement, but a predictive asset for biomanufacturing, product development and supply chain operations. Researchers led by Roger Hart, PhD and Kelvin Lee, PhD from the National Institute for Innovation in Manufacturing Biopharmaceuticals, advocate both leveraging big data within companies and standardizing and sharing key insights across the industry.

This comes from a big data model developed through the MELLODDY Consortium with data from companies like Amgen, Astellas, AstraZeneca, Bayer, Boehringer Ingelheim, GSK, Janssen, Merck KGaA, Novartis and Servier. Researchers pooled decades of proprietary data, including 2.6 billion data points and over 40,000 assays, to enable a predictive drug-activity model.

Their results significantly outperformed those of individual members and enhanced predictivity in most classification or regression tasks, with a median Relative Improvement of proximity to perfection for conformal efficiency exceeding 12%. While individual organizations pursue in-house solutions, participating in larger-scale efforts can benefit from shared learnings and accelerate the digital transformation of the entire industry.

However, achieving these benefits requires significant data management preparation, such as standardizing ontologies, schemas, and holistic data integration, enabling predictive strategies to optimize processes, developing secure shared datasets, ensuring interoperability and real-time data connectivity, and overcoming barriers like regulatory uncertainty and a lack of workers with expertise in both biomanufacturing and data science.

Companies that fail to adopt these capabilities risk falling behind in realizing the full benefits of big data in biopharmaceutical manufacturing.