Purchase this article with an account.
P. Benaglio, E. Brini, C. Rivolta; Challenges of Ultra High Throughput Sequencing in Mutation Detection: A Benchmark Study on an Intronic Repetitive Element of the Retinitis Pigmentosa 11 Gene. Invest. Ophthalmol. Vis. Sci. 2010;51(13):4090.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recently, we discovered a causative mutation for autosomal dominant retinitis pigmentosa (adRP) in a repetitive element (VNTR) nested in an intron of the PRPF31 gene (RP11). Because of the sequence context of this variant, its detection represents a real challenge for both traditional and ultra high throughput sequencing (UHTS). To evaluate UHTS as a routine tool for the detection of human mutations, we analyzed a 31-kb genomic region spanning the PRPF31 gene in an individual carrying this mutation with different UHTS technologies.
We used the latest versions of the three most widespread UHTS platforms: Roche/454 GS FLX Titanium (generating long reads), Illumina/Solexa Genome Analyzer II, and Applied Biosystems SOLiD System 3 (both generating short reads). The template DNA material was a pool of 4 long-range PCRs encompassing the PRPF31 genomic region. All the post-sequencing analyses were performed using standard commercial software.
The performance of the three platforms was assessed in terms of SNP detection. Although we overall obtained similar results (94%, 96% and 96% of known SNPs detected, respectively), we observed that undetected variants were specifically located in regions with critical features for long read chemistry (Roche/454) or for short read alignment (Illumina and AB SOLiD). The intronic mutation associated with adRP was clearly identified by all three technologies. However, no UHTS experiment could reconstruct the exact sequence of the VNTR (i.e. the number of repeated modules in the analyzed individual), which was disclosed only by Sanger sequencing. We also observed that long reads (200-300 bp) could map to the entire repetitive region. Conversely, as expected, most short reads (35-50 bp) had multiple matches and were therefore discarded from the alignment, failing in the end to cover the whole repeat.
Our study demonstrates the possibility of detecting mutations in a rapid and cost-effective manner using UHTS, while identifying at the same time the weak points to be overcome for systematic use of this technology in targeted resequencing of candidate genes or of genomic intervals for disease association studies.
This PDF is available to Subscribers Only