Using a constraint-based regression method for relative quantification of somatic mutations in pyrosequencing...

Area of Science:

• Genomic analysis
• Molecular diagnostics
• Bioinformatics

Background:

• Pyrosequencing Allele Quantification (AQ) is a sensitive, rapid method for detecting somatic mutations in FFPE samples.
• However, Pyrosequencing AQ has limitations including low specificity and difficulty interpreting signals with multiple mutations in a genomic hotspot.
• These drawbacks hinder its clinical utility for complex mutation profiling.

Purpose of the Study:

• To develop a novel algorithm, AdvISER-PYRO-SMQ, to overcome the limitations of standard Pyrosequencing AQ.
• To enhance the detection of multiple somatic mutations within a single pyrosequencing reaction.
• To improve the specificity and signal interpretation for pyrosequencing data.

Main Methods:

• Developed AdvISER-PYRO-SMQ, a constraint-based regression algorithm implemented as an R package.
• Applied AdvISER-PYRO-SMQ to identify 9 distinct NRAS oncogene mutations at codon 61.
• Compared AdvISER-PYRO-SMQ performance against Qiagen's AQ module using pyrosequencing assays.

Main Results:

• AdvISER-PYRO-SMQ demonstrated a lower limit of blank (0%) compared to Qiagen's AQ module (5.1%).
• Both methods achieved similar limits of detection (4.8-5.6%).
• AdvISER-PYRO-SMQ successfully screened for 9 mutations in one reaction, unlike the AQ module limited to single mutations.

Conclusions:

• Constraint-based regression analysis enables robust detection of multiple mutations in hotspot regions, balancing sensitivity and specificity.
• The AdvISER-PYRO-SMQ R package is a versatile tool applicable to various somatic mutations.
• An interactive web application enhances accessibility for research and clinical routine use.