DNA Probe SNP Discrimination: Nearest-Neighbor Melting and the Best Assay Temperature

A two-well DNA chip on a heater: blue targets are a perfect match, orange targets carry one mismatch. Turn the temperature and watch the helices melt in the orange well first. Drag to orbit.

Switch to 2D
Controls

Presets

Parameters

Quality

Camera

Model checks

The science

Duplex stability. The free energy of a DNA duplex is the sum over its nearest-neighbour stacks (each base pair interacts with the next one) plus an initiation term for each end: ΔH = ΣΔHstack + init, ΔS = ΣΔSstack + init + 0.368 (N-1) ln[Na+], ΔG(T) = ΔH - TΔS. The ten stack values and the initiation terms are the unified parameters of SantaLucia (1998, PNAS 95:1460), checked in this build against the published table (ΔH, ΔS and ΔG37 columns, 1 M NaCl); the sodium correction follows the same paper (slope -0.114 kcal/mol per ln[Na+] per phosphate pair).

Binding. Probe P and target T make a duplex D with K = exp(-ΔG/RT) (M-1); the exact equilibrium is the quadratic D = 2PT/(b + sqrt(b2 - 4PT)) with b = P + T + 1/K. The kinetics are mass action with kon = 106 M-1 s-1 (a typical order of magnitude for short oligonucleotides) and koff = kon/K. For equal strand concentrations C the midpoint is Tm = ΔH/(ΔS + R ln(C/2)), and the melting curve there has slope dθ/dT = ΔH/(6 R Tm2).

Single-nucleotide discrimination. A single mismatch adds a penalty ΔΔG that depends on where it sits: weak at the ends (the duplex frays anyway), strong in the centre. Wells A and B hold perfect-match and mismatch targets; at low occupancy the signal ratio is exactly exp(ΔΔG/RT), but near saturation both wells fill up, so the best assay temperature sits between the two melting points. Too cold and the mismatch hybridises; too hot and nothing does.

Simplifications. The mismatch penalty is a parameterised number with an illustrative position profile (0.3 + 0.7 sin) and a temperature-independent ΔΔG, not the mismatch tables of SantaLucia & Hicks (2004) which depend on the base pair; surface effects (probe crowding, electrostatics, the reduced affinity of surface-bound probes), secondary structure, Mg2+ and dangling ends are ignored; the fixed 24-mer is an arbitrary example and the sodium range is that where the empirical correction was fitted. Eight independent checks: two kinetics-versus-equilibrium runs, two melting temperatures against the closed form, the van't Hoff slope, parameter-table consistency, strand symmetry and the low-occupancy discrimination limit.

Loading model...
Loading Blender model...
-
-
Made with 🟠 Blender · Three.js