Skip to main content
. Author manuscript; available in PMC: 2021 Aug 30.
Published in final edited form as: J Chem Inf Model. 2018 Mar 6;58(3):591–604. doi: 10.1021/acs.jcim.7b00496

Figure 4.

Figure 4.

ANN-v train and validate results for 1955 compound data set. Round data points on each end represent compounds with experimental values outside the QRI (outside the range of N-amide standards, RIU < 300, RIU > 1400). On the validate data plot, the gray boxes at each end indicate the prediction cutoff for ∼95% true positive predictions. Data points within each box are true positive predictions while those outside are false negatives. These cutoff values approximate the ±2SE filter range for compounds with predicted values outside the range of standards. The center gray box illustrates the y-range of the ±2 SE filter range which constitutes 14.7% of the overall data range of 1381 RIU.