The new technology is described in The Journal of Molecular Diagnostics. The web-based application produces simulated pyrograms based on user input including the percentage of tumor and normal cells, the wild-type sequence, the dispensation order, and any number of mutant sequences. 

Pyromaker then calculates the relative mutant and wild-type allele percentages and uses these to generate the expected signal at each point in the dispensation sequence. The final result is a virtual trace of the expected pyrogram.

The researchers validated Pyromaker against actual pyrograms containing common mutations in the KRAS gene, which plays an important role in the pathogenesis of a variety of tumors. The actual pyrograms and virtual pyrograms were quantitatively identical for all mutations tested.

They then demonstrated that all codon 12 and 13 single and complex mutations generate unique pyrograms. However, some complex mutations were indistinguishable from single base mutations, indicating that complex mutations may be underreported.

Five solution approaches
Working with two complex pyrograms that were difficult to interpret initially, the researchers identified five approaches to resolve them: Sanger sequencing alone, hypothesis testing with Pyromaker, Pyromaker iterative mutation re-creation, melting curve analysis, and TA cloning with Sanger sequencing.

Senior author, (right) James R. Eshleman, MD, PhD, Professor of Pathology and Oncology, Associate Director, Molecular Diagnostics Laboratory, Johns Hopkins University School of Medicine, explains, “User-directed hypothesis testing allows for generating virtual traces that can be compared to the actual data to clarify ambiguous results from pyrosequencing and the Sanger method.

"Alternatively, Pyromaker can quickly and efficiently test the possibilities that can explain a complicated polysequencing result.” Both strategies were able to successfully identify the complex mutations.