FIGURE 2.

z-score from the initial, confirmation and counter screens. In (A,B), the z-scores of every compound evaluated during the initial screen are plotted. Compounds were from two sub-libraries: one containing compounds with demonstrated biological activity (biologically active library) (A) and the other representing a general small molecule library (generic library) (B). A total of 8,320 compounds were tested in the initial screen, 4,480 of which came from the biologically active library and the remaining 3,840 from the generic library. Representative OD₆₀₀ profiles that either passed (outlined in green) or did not pass (outlined in orange) the z-score threshold of –3 (horizontal gray line) were plotted on the top. The growth curve associated with 2-MBT treatment from the initial screen is plotted in the green box in (A). In (C,D), the z-scores of hits from the initial screen during the confirmation and counter screens are plotted. In (C), growth inhibition in the presence of NO was verified. Compounds with high z-scores allowed growth and were identified as outliers (circle in red). We note that since the confirmation screen only included hits, the z-scores between the primary screen and confirmation screen will not be the same. In (D), growth in the absence of NO was measured. NO-specific growth inhibitors would not inhibit growth during this screen, so compounds with the higher z-scores allowed relatively better growth. A natural break in the data occurs around a z-score of 0 in the counter screen, therefore compounds with z-scores > 0 were identified as growth permissive. Of note, out of 20 compounds that had a positive z-score in the counter screen, 3 were outliers in the confirmation screen (C) and thus were disqualified, and 17 compounds were identified as hits from the screening procedure.