What is Biomedical Data Science and Do We Need an Annual Review of It?

We are pleased to bring you the first volume of the Annual Review of Biomedical Data Science. It spans a range of biological and medical research challenges that are data intensive and focused on the creation of novel methodologies to advance biomedical science discovery. The term “data science” describes expertise associated with taking (usually large) data sets and annotating, cleaning, organizing, storing, and analyzing them for the purposes of extracting knowledge. It merges the disciplines of statistics, computer science, and computational engineering. Many are irritated by the term—all of science depends ultimately on data, and many of the activities listed above sound like engineering (which is about solving problems) and not science (which is about discovery of new knowledge). If “data science” is not about science and the adjective “data” has no particular meaning, why does this term exist? Indeed, the allied fields of informatics have existed for several decades in many forms—medical informatics, clinical informatics, health informatics, bioinformatics, and biomedical informatics—and variants all refer to the development of methods to analyze data, information, and knowledge within the space of biology and medicine. Practitioners of these fields are quick to point out that most if not all of data science falls within the purview of informatics. Informatics is a broad field that includes the social aspects of interacting with data, information, and knowledge; the challenges of human–computer interfaces; and the issues associated with introducing disruptive new computational interventions into systems (like hospitals and laboratories) with existing workflows. So why is the introduction of a new name for the field necessary? The term “data science” has gained recognition, and the widespread comfort with it suggests it serves a useful purpose. Here we offer some observations on the diverse use of the moniker for many activities: