Fig. 2: Identifying sensors in a ligand agnostic library.

a Ligands used in this study. Tanimoto scores (Tan.) are provided comparing each ligand to native ligands, naringenin (Nar) and phoretin (Phlo). b Histograms showing distribution of F-scores for each variant in the library with each tested ligand. Red line denotes the 99th percentile of F-scores. c RNA-Seq fold enrichment data for 16,191 variants which passed filters across nine ligands. Ligands and variants have been clustered via the UPGMA algorithm with a correlation distance metric and a target of 12 clusters (see “Methods”). The different clusters are denoted by the colored bars on the left of the heatmap. aTF function is shown as the log₂(F-score) normalized to wildtype (WT). d Violin plot showing distribution of fold enrichment scores of a 251-member library consisting of the top 40 variants for each ligand (determined by F-scores) after induction with indicated ligand and fluorescence-based sorting. The box represents the interquartile range (IQR). Whiskers extend to 1.5 times the IQR. Center dot represents the median. Red circles indicate scores of variants selected as top 40 for corresponding ligands. e Heatmap showing clonal measurements of fold induction of WT TtgR and variants that were selected as top 3 for each ligand based on data from (d). Right panel consists of non-native ligands, and the left panel consists of the two native ligands. A score ceiling of 20 was imposed for visualization. Black circles within heatmap cells reflect the coefficient of variation (CV). f Bar plot of clonal measurements of variants with a specificity switch for the native ligands. Bars represent the mean fold induction ± SD (n = 3 independent replicates). Source data are provided as a Source Data file.