FIG. 3.

Summary of observed hybridization kinetics. (a) Observed k_Hyb value (model H3) for 96 targets at 37°C and 55°C. 4 A/T targets were excluded from this because they were A/T rich and did not stably bind to their probes at 55°C. (b) Most reactions did not reach completion, instead saturating at between 60% and 100% yield. Yield is determined based on positive control experiments wherein target and probe are thermally annealed (Supplementary Section 1). We modeled incompleteness of hybridization as a “bad fraction” of probes that becomes kinetically trapped at a high fluorescence state. The best-fit bad fractions for the 96 targets at 37°C and 55°C are plotted here. (c) There appears to be no correlation between k_Hyb and asymptotic yield. Blue and red dots show experiments at 37°C and at 55°C, respectively. (d) Systematically designed target/probe sequences included 13 clusters, each comprising 3 targets. Within each cluster, the target sequences were shifted such that predicted secondary structure is present (1) near the 5′ end (plus symbols), (2) near the middle (circles), or (3) near the 3′ end of the target (triangles). In 8 out of 13 clusters, the target with structure in the middle was the slowest. (e) Because secondary structures are known to slow down kinetics [23], we examined the target/probe pairs in which both the target and the probe had predicted ensemble (partition function) standard free energy (ΔG°_pf) of greater than −3 kcal/mol at the experimental hybridization temperature, indicating minimal structure. At 37°C and at 55°C, 29 of the 100 reactions and 61 out of 96 reactions satisfied this criterion, respectively. These reactions all have k_Hyb ≥ 10⁶ M⁻¹s⁻¹, but k_Hyb values range more than 1 order of magnitude. (f) Minimal-structure targets exhibit significant variability of bad fraction, ranging between 0% and 25%.