“…All the computational methods discussed here, such as MD simulation or in silico docking, require high-resolution 3D structure of the protein under investigation. While murine Abcb1 has been crystallized in an apo state (Aller et al, 2009), and sufficient homology models can be built for its ATP-bound conformation (Globisch et al, 2008;O'Mara & Tieleman, 2007;Pajeva, Globisch, & Wiese, 2009), structural or homology models of human ABCG2 are insufficient for in silico studies because of the low-resolution structure information (cryoelectron microscopy, >5 Å ) (Rosenberg et al, 2010) and the very low (<20%) sequence similarity of the TMD with any existing ABC structure that could be applied as a possible template. s0110 5.1.…”