Figure 1: Outline of screen. Fragment screen performed first was a structure-based virtual screen on small molecules. Product of that screen was information, motifs. Filtration was based on the output of the first screen and acted on a larger, lead-like database. Filtered output was subjected to a second screen, again structure-based virtual screening. High-ranking molecules from the second screen were subjected to MM-GBSA analysis to generate a list of “hits.”