An infrastructure for qualified data sharing and team science in late-stage translational spinal cord injury research

Abstract

The complex and heterogeneous nature of spinal cord injury has limited translational bench-to-bedside results. The wide variety of data, including injury parameters, biochemical, histological, and behavioral outcome measures represent a ‘big data’ problem, calling for modern data science solutions. There are some instances in which SCI researchers collect sensitive data that needs to remain private, such as datasets designed to meet regulatory approval, sensitive intellectual property, and non-human primate studies. For these types of data, we have developed a Private Data Commons for SCI (PDC-SCI). Our objective is to give an overview of this novel data commons, describing how this type of commons works, how it can benefit the research community, and the cases in which it would be most useful. This private infrastructure is ideal for multi-lab transdisciplinary studies that require a well-organized, scalable data commons for rapid data sharing within a closed, distributed team. As a use-case for the PDC-SCI, we demonstrate the VA Gordon Mansfield SCI Consortium, in which multimodal data from behavior, biomechanics of injury, hospital records, imaging, and histology are integrated, shared, and analyzed to facilitate insights and knowledge discovery.