Abstract
A major challenge facing the biomedical community is creating and sustaining high-quality research environments. A literature search identified five common themes underlying biomedical research environments comprising collaboration, data access, user-led innovation, data provenance and a deep commitment to public and scientific benefit. Club theory is used to develop a model describing social structures that underpin these themes. It is argued that collaboration underlies impactful science and that collaboration is hindered by high transaction costs. This, combined with poorly defined property rights surrounding publicly funded data, limits the ability of data markets to operate efficiently. Although the science community is best placed to provide solutions for these issues, incentivization by funding agencies to increase the benefits of collaboration and reduce uncoordinated activity will be an accelerator. Given the complexity of emerging datasets and the collaborations needed to exploit them, trust-by-design solutions are suggested. The underlying motivational 'glue' that holds this activity together is the aesthetic and ethical value base underlying good science. The model has implications for data-driven science more generally. As biomedical science in the Global South develops, there is an opportunity to address foundational structural issues prospectively rather than inherit unwanted constraints of current practice.