Accurate variant detection across non-amplified and whole genome amplified DNA using targeted next generation sequencing.

Authors:
Abdou ElSharawy, Jason Warner, Jeff Olson, Michael Forster, Markus B Schilhabel, Darren R Link, Stefan Rose-John, Stefan Schreiber, Philip Rosenstiel, James Brayer, Andre Franke
Year of publication:
2012
Volume:
13
Issue:
-
Issn:
1471-2164
Journal title abbreviated:
BMC GENOMICS
Journal title long:
BMC genomics
Impact factor:
3.729
Abstract: 
Our described PCR-based sample multiplex approach and the ability to use WGA material for NGS may enable researchers to perform deep resequencing studies and explore variants at very low frequencies and cost.The results demonstrated comparable sequence performance between non-amplified and preamplified samples and between different indexing strategies [sequence specificity of 66.0% ± 3.4%, uniformity (coverage at 0.2× of the mean) of 85.6% ± 0.6%]. The average genotype concordance maintained across all the samples was 99.5% ± 0.4%, regardless of sample type or pooling strategy. We did not detect any errors in the Mendelian patterns of inheritance of genotypes between the parents and offspring within each trio. We also demonstrated the ability to detect minor allele frequencies within the pooled samples that conform to predicted models.Many hypothesis-driven genetic studies require the ability to comprehensively and efficiently target specific regions of the genome to detect sequence variations. Often, sample availability is limited requiring the use of whole genome amplification (WGA). We evaluated a high-throughput microdroplet-based PCR approach in combination with next generation sequencing (NGS) to target 384 discrete exons from 373 genes involved in cancer. In our evaluation, we compared the performance of six non-amplified gDNA samples from two HapMap family trios. Three of these samples were also preamplified by WGA and evaluated. We tested sample pooling or multiplexing strategies at different stages of the tested targeted NGS (T-NGS) workflow.