We have used crystal structures and molecular modeling to evaluate inhibitor binding modes and design a series of compounds to take advantage of a new, cryptic, hydrophobic sub-pocket. This is a classical SBDD approach to improving enzyme/inhibitor interactions.