This paper examines the ways in which the plant science community, in collaboration with other communities of researchers in biomedicine, has re-organised its research practices and infrastructures to enable the intelligent re-use of data over the last 25 years; and how the development of infrastructures and forms of governance aimed at circulating molecular data has affected the ways in which biologists today think about what it means and what it takes to transform large quantities of data into scientific insights. I first examine how the introduction of genome sequencing in the 1990s provided an excellent opportunity for biologists to start re-structuring their means of exchanging molecular data, and devise institutional mechanisms, infrastructure and relevant expertise that would facilitate data-intensive discovery. These initiatives include societies like GARNet and data curation initiatives such as the Plant Ontology, and are very heavily structured around the study of one plant species, Arabidopsis thaliana, as a reference organism. I then discuss how these initiatives have had two seemingly contrasting effects on contemporary plant science. On the one hand, they have boosted interest in so-called Open Data initiatives particularly in the UK, thus fostering attention to the dissemination and re-use of data well beyond the molecular level and single species (e.g. data about the physiology, metabolism, morphology and ecology of a large variety of plants). On the other hand, they have put molecular data, and particularly high-throughput data from the genomics, proteomics and metabolomics of Arabidopsis, at the centre of data-intensive research, thus transforming one specific type of data on one specific type of organism into an obligatory passage point to study the natural variation, biodiversity and ecology of plants.