11.2.3. High Resolution Melting (HRM) analysis

Double stranded DNA can disassociate (or melt) into two single strands upon heating, and can re-associate (or hybridize) upon cooling, in a highly predictable fashion. This fundamental property of nucleic acids underpins all nucleic acid technologies. The principal parameters governing disassociation/hybridization are the length and composition of the DNA, the temperature and the salt concentration of the solution. Work in the 1950s demonstrated that the G-C pairing, with three hydrogen bonds, gave higher thermal stability than the A-T pairing, which has only two such bonds (Marmur and Doty, 1959). This made it possible to predict the temperature at which a DNA molecule would melt (Tm) from its length and base pair composition (Marmur and Doty, 1962). The discovery of DNA binding dyes such as SYBR-green, that fluoresce only when intercalated with double stranded but not single stranded molecules, provided a practical method to quantify the melting process based on a reduction in fluorescence during gradual heating, as the two DNA strands separated. This fluorescence-based detection was integrated with real-time PCR thermocyclers that can very precisely control the temperature of a DNA sample and collect fluorescence data between 10 and 200 times per °C, providing high-resolution melting curves that can distinguish single base pair differences between two PCR products (Wittwer et al., 2003). This makes it possible to use High Resolution Melting (HRM) analysis to analyse the composition of mixed samples, i.e. samples containing two or more genetic variants of the same region, by comparing the melting curve of the mixed sample with those of the individual variants. 

HRM analysis is a versatile method that can be applied to any sample that contains double stranded DNA, including cDNA or PCR products. The flexibility of HRM analysis has led to a diverse array of applications including pre-sequence screening, Single Nucleotide Polymorphism (SNP) typing, methylation analysis, microsatelite or Simple Sequence Repeat (SSR) marker screening (Arthofer et al., 2011) and copy number quantification. Several of these techniques are covered in the BEEBOOK paper on molecular techniques (Evans et al., 2013). Such applications of HRM are also relevant to virology, and the first record of a virological use of HRM analysis was to strain type West Nile virus (Papin et al., 2004). Recently, HRM analysis has been used to monitor the relationship between varroa infestation and virus diversity (Martin et al., 2012).

Many standard real-time PCR machines can be used for HRM analyses. Often an upgrade of the software package and the running of a calibration plate is all that is required to enable a real-time PCR machine to run HRM analyses. Since HRM is a highly technical and sensitive procedure that integrates reaction biochemistry with machinery and analysis software, the best advice is to follow the protocol, reagents and incubation profile recommended by manufacturer. The basic procedure is as follows:

1. Amplification:     Amplify your chosen fragment from your experimental samples and cloned controls, using specific HRM reagents containing a saturating DNA DNA intercalating dye and the recommended incubation profile. 

2. Replicates: Use a minimum of three technical replicates for each sample. The replicate melting profiles will be averaged and used to assess whether the sample is distinct from other samples/controls.

3. HRM: Immediately after amplification the PCR products are subjected to a high-resolution melting step, within the same tube, during which the decrease in fluorescence due to the transition of the DNA from double- to single-stranded shape is monitored.

4. Analysis: The melting curve of the experimental sample, containing a mixture of different variants, is compared to the melting curves of pure, cloned versions of each of the individual variants.  


  • Simple; fast; flexible; cheap; sensitive; specific; low contamination risk.


  • Requires individual melting curves of (cloned) variants.
  • Cannot identify nature of novel variants.
  • Limited quantification of variants.
  • Limited capacity to resolve complex mixture.
  • Limited to very short genome fragments.