Keywords: data integration, big data, machine learning, biomarker, electronic record
The changing technology landscape is having a profound impact on the regulatory sciences in the biopharmaceutical area. There are a multitude of emerging technologies generating a high volume of data that promise to transform various aspects of product discovery, development, and regulatory approval. Some notable examples are next generation sequencing based approaches for target discovery, diagnosis, and subgroup identification; “omics” based biomarker development, high throughput toxicological screening; electronic medical record based knowledge discovery and surveillance; among others. However, the ability to rapidly generate a huge amount data also poses significant challenges for statisticians and other data scientists. There are pressing needs for data standardization, data storage, data processing and analysis, and data integration. Integrative analysis on a large scale is especially challenging given the heterogeneous nature of data generating technologies. To derive useful knowledge for regulatory purpose, one need to draw on various branches of statistics and information science; including Bayesian approaches, high dimensional methods, machine learning; and utilize novel computing infrastructure. In this round table, we will discuss various aspects of integrative statistical analysis for biopharmaceutical applications with an emphasis on “big data”.