Fig. 5: PNRSS sensing of epinephrine, norepinephrine and isoprenaline.

a The reaction mechanism. Norepinephrine, epinephrine and isoprenaline all contain a 1,2-benzenediol moiety, capable of binding a PBA. b Representative events of epinephrine, norepinephrine or isoprenaline binding. The events were low-pass Butterworth filtered with a cut-off frequency of 100 Hz (Supplementary Fig. 44). All events appear negative going (\({I}_{{{{{\mathrm{b}}}}}} \; < \; {I}_{{{{{\mathrm{p}}}}}}\)). c, d A representative trace containing norepinephrine-, epinephrine- or isoprenaline-binding events. The trace was Butterworth filter separated into the low-pass (c) and the high-pass (d) portion. The cut-off frequency is 100 Hz (Supplementary Fig. 44). e A confusion matrix was generated based on 1455 events fed into a SVC model (Supplementary Fig. 45). f The scatter plot of low-pass (Lp) and high-pass (Hp) amplitude SD from PNRSS sensing of a mixture of catecholamine events (Supplementary Fig. 46). Results of 150 events are included. The decision boundary, which separates the scatter plot to green (norepinephrine), pink (epinephrine) and orange (isoprenaline) colour-coded regions, was determined by a machine learning algorithm (Supplementary Fig. 45).