Extended Data Fig. 4: Serial dilutions and standard curves of 3 clinical specimens.

(a) Standard curve relating Ct value to spiked virus concentration, averaged across 3 replicates on a standard method (QuantStudio). This standard curve was used to calculate concentrations of the clinical specimens. Standard deviation error bars present but not visible due to small range. Slope was -3.22. (b) N1 Ct values for serial dilutions of three different clinical specimens. Limit of detection (lowest concentration with 3/3 replicates detected below threshold) of less than 10⁴ copies/mL was obtained for all 3 specimens. Data are presented as mean values + /- SD. Threshold was set at 45 cycles, and samples that were not detected during amplification were plotted as having Ct values of 46. Concentrations are approximated in legend; sample 1 exact starting concentration was 273,771 copies/mL, sample 2 exact starting concentration was 1,521,025 copies/mL, and sample 3 exact starting concentration was 7,470,693 copies/mL. Symbols indicate statistical significance for the following comparisons: □ against ~106 copies/mL, ■ against ~105 copies/mL, △ against ~104 copies/mL, ▲ against ~5000 copies/mL, ☆ against ~2500 copies/mL, ★ against ~1250 copies/mL, ○ against ~625 copies/mL, ● against ~312 copies/mL. (c-e) Three separate standard curves relating Ct value to viral concentration from dilution series from three different clinical specimens. Data shown are averaged across 3 replicates run on the plasmonic instrument, presented as mean values + /- SD. The slopes of the three separate dilution series were −3.40 (d), −3.56 (e), and −3.91 (c), corresponding to acceptable amplification efficiencies (that is high efficiencies of >90% for (d) and (e), and still >80% for (c)), consistent with the demonstrated high sensitivity and specificity of the method on clinical specimens.