High-resolution melting of DNA is a simple solution for genotyping, mutation scanning and sequence matching. The melting profile of a PCR product depends on its GC content, length, sequence and heterozygosity and is best monitored with saturating dyes that fluoresce in the presence of double-stranded DNA. Genotyping of most variants is possible by the melting temperature of the PCR products, while all variants can be genotyped with unlabeled probes. Mutation scanning and sequence matching depend on sequence differences that result in heteroduplexes that change the shape of the melting curve. High-resolution DNA melting has several advantages over other genotyping and scanning methods, including an inexpensive closed tube format that is homogenous, accurate and rapid. Owing to its simplicity and speed, the method is a good fit for personalized medicine as a rapid, inexpensive method to predict therapeutic response.

No abstract

No abstract

Melting curve prediction of PCR products is limited to perfectly complementary strands. Multiple domains are calculated by recursive nearest neighbor thermodynamics. However, the melting curve of an amplicon containing a heterozygous single-nucleotide variant (SNV) after PCR is the composite of four duplexes: two matched homoduplexes and two mismatched heteroduplexes. To better predict the shape of composite heterozygote melting curves, 52 experimental curves were compared with brute force in silico predictions varying two parameters simultaneously: the relative contribution of heteroduplex products and an ionic scaling factor for mismatched tetrads. Heteroduplex products contributed 25.7 ± 6.7% to the composite melting curve, varying from 23%-28% for different SNV classes. The effect of ions on mismatch tetrads scaled to 76%-96% of normal (depending on SNV class) and averaged 88 ± 16.4%. Based on uMelt (www.dna.utah.edu/umelt/umelt.html) with an expanded nearest neighbor thermodynamic set that includes mismatched base pairs, uMelt HETS calculates helicity as a function of temperature for homoduplex and heteroduplex products, as well as the composite curve expected from heterozygotes. It is an interactive Web tool for efficient genotyping design, heterozygote melting curve prediction, and quality control of melting curve experiments. The application was developed in Actionscript and can be found online at http://www.dna.utah.edu/hets/.

Fluorescent high-resolution DNA melting analysis is a robust method of genotyping and mutation scanning. However, removing background fluorescence is important for accurate classification and to correctly display helicity. Linear baseline extrapolation, commonly used with absorbance, often fails at low temperatures when fluorescence is used. A new quantum method of background removal based on the inherent decrease of fluorescence with temperature is described. Absorbance and fluorescence melting curves were compared using synthetic targets including hairpins, unlabeled probes, and a 50 bp duplex. In addition, the quantum method was compared to a previously described exponential method for analysis of genotyping data produced after polymerase chain reaction (PCR), including those from small amplicons, unlabeled probes, and snapback primers. The quantum method best matched absorbance data and predicted helicity, with the exponential method displaying low-temperature bulges and domain artifacts that can lead to incorrect genotyping. When two melting domains were widely separated, quantum analysis produced a flat baseline between domains, while exponential analysis was temperature-dependent. Both methods have little effect on the melting temperature (Tm) although some differences were significant (hairpin Tm values increased 0.7 °C by the quantum method and decreased 1.5 °C by exponential method, p = 0.01). However, peak heights on derivative plots were strongly algorithm-dependent, with exponential analysis enhancing low-temperature peaks while dampening high-temperature peaks. Quantum-analyzed fluorescence curves were a better match to absorbance data in terms of shape, area, and peak height compared to other methods, indicating that DNA helicity is best approximated by the quantum method.