Background: Denaturing HPLC (DHPLC) is a semi-automated method for detecting unknown DNA sequence variants. The sensitivity of the method is dependent on the temperature at which the analysis is undertaken, the selection of which is dependent on operator experience. To circumvent this, software has been developed for predicting the optimal temperature for DHPLC analysis. We examined the utility of this software.
Methods: To maximize the relevance of our data for other investigators, we have screened 42 different amplimers from CFTR, TSC1, and TSC2. The samples consisted of 103 unique sequence heterozygotes and 126 wild-type homozygous controls.
Results: At the temperature recommended by the software, 96% (99 of 103) of heterozygotes and all of the wild-type controls were correctly classified. This compares favorably with sensitivities of 85% for single-stranded conformation polymorphism and 82% for gel-based heteroduplex analyses of the same fragments.
Conclusions: Software-optimized DHPLC is a highly sensitive method for mutation detection. However, where sensitivity >96% is required, our data suggest that in addition to the recommended temperature, fragments should also be run at the recommended temperature plus 2 degrees C.