A Digital Twin (DT) is a dynamic and continuously updated digital representation of a physical system, powered and driven by data. The Digital Thread (DTH), a bidirectional flow of data, is the core element of a DT, and enables real-time synchronisation between physical and digital entities. The value generated by a DT depends on its ability to accurately represent the physical entity, monitor and predict its evolution over time, and interact with it in order to manage and improve it. This study explores the role of data in Digital Twins (DTs). Initially, two frameworks are proposed. The first is the Digital Twin Value Chain, inspired by the Big Data Value Chain, which models the phases of DT projects and highlights the processes through which data are transformed into informational and strategic value. The second is a modelling of DTH. It captures the central role of data flows in driving value creation through the Functions of the DT. Subsequently, the focus is on the integration of heterogeneous DTs and on their interoperability. To this aim, a third framework for a network of interconnected DTs is proposed, structured according to the principles of Data Mesh. The three proposed frameworks were developed through a qualitative and inductive approach, based on an exploratory analysis of the literature and the observation of critical issues emerging in existing models. They provide valuable insights for future implementations and developments. Moreover, the integration of Generative AI technologies is shown to promote interoperability among DTs, and enhance the development of increasingly advanced systems.