<p>Despite the growing science-technology-industry interconnectedness, research on the interactions remains significantly under-developed. It is well known that organizations, as the primary agents of innovation, play a crucial role in fostering collaborative innovation across science, technology, and industry. Therefore, this paper proposes a percolation framework from the perspective of organization cooperation to model the science-technology-industry interactions. To assess the feasibility and effectiveness of our framework, a three-layer interacting network of authors’ affiliations, patent applicants, and drug manufacturers is constructed on the basis of STInt dataset. Extensive empirical analysis of the network reveals significant differences in the strength of intra- and inter-layer interactions as follows. (1) The science layer exhibits active intra-layer interactions, but relies heavily on cross-layer cooperation for development. (2) The technology layer acts as a key bottleneck in the science-technology-industry interactions. While its interaction with the industry layer is relatively active, its cross-layer cooperation mechanism needs to be strengthened. (3) The industry layer demonstrates a high degree of interaction, driving the activity of both the science and technology layers and playing a pivotal role in facilitating cross-layer interactions. While the proposed percolation framework is general in design, the empirical conclusions reflect characteristics specific to the pharmaceutical sector. Therefore, the findings should be interpreted within this domain context, and future studies may apply the framework to other industries to assess its broader applicability.</p>