A long and winding road: Culture change on data sharing in exposomics

Abstract

Data sharing requires cooperation from data generators (e.g., epidemiologists, lab investigators) and data users (e.g., epidemiologists, biostatisticians, computer scientists). Data generation and data use in human exposome studies require significant but different skill sets and are separated temporally in many cases. Sharing will require maintaining a history of data generation and a system to address the concerns of data generators around credit for conducting rigorous work (e.g., authorship). Sharing also requires addressing the needs of data users to facilitate harmonization, searchability and QA/QC of data. We present these issues from the perspectives of data generators and data users and include the special case of real-world data (e.g., electronic health records). We conclude with recommendations to address how to better promote data sharing in exposomics through authorship, cost recovery and addressing ethical issues.