Thus, in silico generation of homology models is typically necessary to capture the structural specificities of human membrane proteins, and this is not different for the ABC superfamily. In addition, it is important to note that of these ~670 membrane protein structures (among a database of approximately 125,000 structures) only 37 of these are ABC proteins (~0.5% of membrane proteins or ~0.03% of all structurally characterized proteins), and most have been deposited in recent years; further, only five full-length human ABC proteins have been structurally characterized to date [32, 33].
In general, for diseases caused by missense mutations to certain members of the ABC superfamily, the first step towards drug discovery and rational drug design is to model these side-chain variants onto the tertiary protein structure and subsequently conduct molecular dynamics (MD) simulations to "relax" the biomolecular system so that it reaches its lowest energy conformation within the biological constraints of a phospholipid bilayer environment [14, 52, 53].
Another ABC protein, CFTR (ABCC7), is the most unique member of the ABC superfamily. Cloning of the CFTR gene in conjunction with the discovery of CFTR as a phospho-regulated chloride channel was seminal in the development of high-throughput, target-based initiatives for Cystic Fibrosis drug discovery [15,73-75].