Building Process Understanding for Vaccine Manufacturing Using Data Mining

ABSTRACT

The production of vaccines is a complex biological process, with long cycle times and high variation in raw materials, growth rates, and test methods. Hundreds of variables are monitored for every batch of vaccine produced; however, the relationships between product quality and process variables are difficult to quantify. We describe how mining historical process data using random forests and partial least squares techniques enabled us to identify the drivers of variability in bulk vaccine yield and to implement new controls which significantly reduced the variation.

KEYWORDS:

• biotechnology
• CART (classification and regression trees)
• data mining
• exploratory data analysis
• hypothesis generation
• multivariate data analysis
• PLS (partial least squares regression)
• random forests
• vaccines

ACKNOWLEDGMENTS

The success of this project depended on the dedication and collaboration of a large number of experts in manufacturing, biology, engineering, bioassay, Six Sigma, statistics, and chemometrics. In particular, we wish to acknowledge the persistence, skill, and leadership of Laura Kasprow in this effort.