Fig. 3: Measuring and predicting RNA-regulatory activity.
From: Designing small molecules targeting a cryptic RNA binding site through base displacement
a, Box-and-whisker plot of the OD600-normalized relative fluorescence units (RFU) of E. coli cells with and without reporter plasmid (env8–GFPuv) and ligand (MeCbl and Cbl 4) added. Data are shown from independent experiments (n = 4). b, Fold repression (defined as the ratio of (−)-Cbl RFU and (+)-Cbl RFU values)33 for our env8 reporter system in the presence of Cbls 1–44 shown as the mean and s.d. from biological replicates (n = 4 except for data from 6, where n = 3). c, Plot comparing the log-transformed fold repression and KD data. d, Locations of the training and test set from the Q2-focused modeling in two-dimensional chemical space constructed from PC1 and PC2 of the whole dataset. e, Measured fold repression values plotted with the value predicted by the Q2-focused model. f, Overview of our lead compound screen from a 513-compound alkyne library using our function-based models. g, Simplified chemical structures of representative derivatives that our function-based models predict to be strong (54–57), moderate (58 and 59) and weak (60–62) repressors. h, Measured fold repression values plotted with the value predicted by our function-based models. The experimental data are presented as the mean and s.d. from biological replicates (n = 4, except for data from 58, 59, 61 and 62, where n = 3) and the predicted data are shown as the mean and s.d. from independent predictions (n = 3) from our three models (that is, baseline, R2-focused and Q2-focused).
