Fig. 3: Biosensors from DRIVER exhibit high affinities and selectivities.
a A high-diversity DNA library of potential RNA biosensors are used as inputs to DRIVER. For each round of selection, the library is transcribed to RNA. For rounds with ligands present (right yellow shaded workflow arrow) the cDNA corresponding to uncleaved product is amplified with a PCR primer specific to the prefix attached to the input to the current round. For selection rounds without ligand present (left blue shaded workflow arrow), the cleaved product is prepended with a new prefix using the regeneration method shown in Fig. 2 and is amplified using a PCR primer specific to this prefix. b Cleavage fractions in the presence and absence of the T1b ligand mixture for a DRIVER-enriched library at round 74 (N = 4328; at least 30 reads/sequence in each condition). Bottom panel shows the standard error of the fold change in cleavage fractions. Dotted lines delineate the region where a multiple-hypothesis test (α = 1/N) would reject the null hypothesis of non-switching; dashed line shows where the fold change of cleavage is 3×; red dots indicate sequences with strong (>3×), significant switching; green crosses indicate validated biosensors that were first identified from this analysis. c SPR binding dose–response curves of select biosensors. Diamonds mark the equilibrium dissociation constants KD, as mean and s.e.m. of n = 3 biologically independent experiments (except for Gard-547, as noted in “Methods”). d Comparison of equilibrium KD and CleaveSeq EC50 for select biosensors. KD are shown as mean and error bars in s.e.m of at least n = 3 biologically independent experiments. EC50s are based on triplicate assays over the ligand and concentrations shown in (e). e Fold change of cleavage fractions for select biosensors measured via CleaveSeq (Supplementary Data 3). Columns represent different biosensor sequences, grouped by ligand. Rows represent ligand concentrations used in the assay. The color represents the fold change of cleavage between the – and + ligand conditions. See also Supplementary Figure 6. f NGS determined the relative abundance of select biosensors during selection rounds immediately prior to their discovery. Data are presented as estimates of the binomial proportion ± standard deviation. The legend indicates the average enrichment/round and extrapolated round 0 fractions based on exponential fits to these data. Source data are provided as a Source Data file.
