Air and blood flow in a set of deformable conduits. Nowadays, computational models of biofluid flow are based on zoomed domains reconstructed from medical image processing. Such modeling is already very useful in medical practice. However it splits the domain of interest from the remaining parts of the network. Most often, crude boundary conditions are used (stress free outlet BCs). Moreover, the living system corresponds to a frosen state, although physiological flows interact with cell lining the interface between fluid and solid. Therefore, computational models of flow in normal and damaged bioconduits require couplings. The talk will illustrate cases for which the nanoscale must be incorporate for future research.