Deriving Insight at the Speed of Thought: Advancing Medicine Production Through an Effective Data Analytics Methodology

by Lisa J. Graham, PhD, PE
Volume 16, Issue 1 (Spring 2017)

In a world already awash with technology, life sciences companies are racing to add more automation and data sources, while ironically often spending less time focused on process improvements. In some cases, these two opposing actions can still produce positive results by: (a) reducing manual labor to minimize data translation errors; (b) adding sensors to gather a new kind of data about a protein or a process; or (c) implementing high-throughput techniques for biopharmaceutical development. But what about those situations where collecting new data is not so positive? Does it really make sense to run experiments without the full benefits of accessing accumulated data or gathering new data? Or to proceed without the insights gained from a colleague down the hall or at another site working on a related project? The difficulty in realizing these potential data analytics benefits often arises because more sensors tend to produce large, complex datasets with multivariate interactions. Further, the inherently complex nature of these datasets makes extraction of meaningful and relevant information a challenging task. This is where a streamlined data analytics methodology can help by providing the foundation to realize the benefits from all of this new data. This article illustrates how a comprehensive data analytics methodology can be used to develop insight into life sciences lab and production data, leading to improved operations. The focus is on sharing lessons learned from recent pharmaceutical case studies to illustrate how to drive innovation through use of a data analytics methodology. These case studies provide detailed, data-driven examples illustrating how to utilize a data analytics methodology to uncover important issues related to pharmaceutical development...

Graham LJ. Deriving insight at the speed of thought: advancing medicine production through an effective data analytics methodology. BioProcess J, 2017; 16(1): 32–41.

Posted online May 8, 2017.