Molecular inversion probe based sequence analysis of 108 genes associated with non-syndromic inherited retinal disease in 4,000 probands

Abstract

Purpose : Inherited retinal diseases (IRDs) are clinically and genetically very heterogeneous, as ~125 genes have been associated with non-syndromic IRDs. The purpose of this study was to develop a flexible, comprehensive and cost effective sequencing procedure for IRDs.

Methods : In total, 6,200 molecular inversion probes (MIPs) were designed to capture ~1600 protein-coding exons and flanking intronic sequences of 108 IRD-associated genes published up to October 2013. About 4,000 probands with non-syndromic IRDs were ascertained by partners of the European Retinal Disease Consortium (ERDC). The most prevalent phenotypes were retinitis pigmentosa, cone-rod dystrophy, and Leber congenital amaurosis. To analyze the performance of the MIPs, three tests were performed. The captured targets were sequenced on a NextSeq500 Illumina sequencer and the data were analyzed using an in-house pipeline to find the causal genetic variants. The over- and poor-performing probes were rebalanced and used to capture the genomic targets in 4,000 individuals.

Results : In pools of 120 samples the average coverage per probe was ~500X, where 95.6% of the probes were covered >10X. Probes targeting 272 (4.4%) regions were either poorly covered (2.3%) or not covered (2.1%). Analysis was completed for 2,500/4,000 samples. Sanger validation was performed for a representative set of variants identified in 290 probands from Nijmegen. Material costs (including MIPs synthesis, library preparations and sequencing costs) for sequencing 108 IRD genes, was € 65 per sample, which makes it very cost-effective. We identified the causal variants in 59% (172/290) of the cases, 36 of 118 unsolved cases (i.e. 12% of total) carry one likely pathogenic variant in an autosomal recessive gene.

Conclusions : Taken into consideration that the Nijmegen cohort was previously prescreened using various genotyping methods, the corrected yield would be ~71%. At 1/10^th of WES cost, this efficiency is equal or superior to other published gene-panel (36 - 62%) or WES-based (49 - 66%) sequence analysis. Based on preliminary results, we estimate to identify causal variants in at least 2,000 IRD probands.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.