Using a constraint-based regression method for relative quantification of somatic mutations in pyrosequencing signals: a case for NRAS analysis. Issue 1 (December 2016)

• Record Type: Journal Article
• Authors: Ambroise, Jerome; Badir, Jamal; Nienhaus, Louise; Robert, Annie; Dekairelle, Anne-France; Gala, Jean-Luc
• Is Part Of: Algorithms for molecular biology. Volume 11:Issue 1(2016)
• Page Start: 1
• Page End: 10
• Publication Date: 2016-12
• Subjects: Pyrosequencing -- somatic -- AdvISER-PYRO-SMQ -- Sparse representation; Molecular biology -- Mathematical models -- Periodicals; Algorithms -- Periodicals; Bioinformatics -- Periodicals; 572.8015118
• DOI: 10.1186/s13015-016-0086-4
• Languages: English
• ISSNs: 1748-7188

Abstract

Abstract Background Pyrosequencing Allele Quantification (AQ) is a cost-effective DNA sequencing method that can be used for detecting somatic mutations in formalin-fixed paraffin-embedded (FFPE) samples. The method displays a low turnaround time and a high sensitivity. Pyrosequencing suffers however from two main drawbacks including (i) low specificity and (ii) difficult signal interpretation when multiple mutations are reported in a hotspot genomic region. Results Using a constraint-based regression method, the new AdvISER-PYRO-SMQ algorithm was developed in the current study and implemented into an R package. As a proof-of-concept, AdvISER-PYRO-SMQ was used to identify a set of 9 distinct point mutations affecting codon 61 of theNRAS oncogene. In parallel, a pyrosequencing assay using the Qiagen software and its AQ module was used to assess selectively the presence of a single point mutation (NRAS $$c.182A>G$$ c . 182 A > G - Q61R-1) among the set of codon 61 mutations, and to analyze related pyrosequencing signals. AdvISER-PYRO-SMQ produced a lower limit of blank (0 %) than the AQ module of Qiagen software (5.1 %) and similar limit of detection were obtained for both software (5.6 vs 4.8 %). AdvISER-PYRO-SMQ was able to screen for the presence of 9 distinct mutations with a single pyrosequencing reaction whereas the AQ module was limited to screen a single mutation per reaction. Conclusion Using a constraint-based regression method enables to analyze pyrosequencing signal and to detect multiple mutations within a hotspot genomic region with an optimal compromise between sensitivity and specificity. The AdvISER-PYRO-SMQ R package provides a generic tool which can be applied on a wide range of somatic mutations. Its implementation in a Shiny web interactive application (available athttps://ucl-irec-ctma.shinyapps.io/Pyrosequencing-NRAS-61/ ) enables its use in research or clinical routine applications.